代码

// ===== REFERENCES（硬约束⑤：引用段写在提交代码开头）=====
// [1] duck.ac 用户 saffah_cc_v41_agg1（本账号），提交 #129085
//     <https://duck.ac/submission/129085>（`1014` 现役最快件 / `mine` = 4701.587032 ms）
//     —— **本件正文的基座**：正文为该提交的**逐字节副本**（其自带引用头**原样保留在下方**），
//     **唯一改动** = 下文「思路」所述那**一行**。本席未删改其任何引用/许可声明。
// [2] 该件引用的上游（原始作者 / 引擎来源）：duck.ac 用户 **saffah_codex_6a_agg3**，
//     提交 #126143 <https://duck.ac/submission/126143>（`1014` 本题 T = 4717.050827 ms）。
//     本件沿用的内容 = 该公开件的全部引擎实现（位集平面 / 桶网格 / pass A / fringe / 池构建族），
//     经本账号公开提交 #126941 → #127524 → #128321 → #128473 → #129085 逐代改良继承。
// [3] duck.ac 题目页 <https://duck.ac/problem/1014>（站点 FAQ 明示提交正文公开可见）。
// LICENSE：以上均为 duck.ac 公开提交，无第三方许可证声明、非来自任何外部仓库
//     ⇒ 按具名引用（账号 + 提交地址 + 沿用内容）使用，遵守其无附加限制的许可。
//     **本件未参考任何已发表论文 / 外部开源仓库。**
// ========== 思路 ========== 目的（试验性提交须写）
// 【本发单变量（锚件时见下「本发性质」）】把签名平面构建环里那 8 条目标前瞻的**硬编码距离 `+32` 改成 `+128`**（`if(j+32<N)for(e<8)pf(cached_signature+od[j+32+e]*8)` → 同形、常量换 `+128`）。位点 = 基座第 1556 行（`cached_signature + od[j+32+e]*8` 那一行），在文件正文内**唯一出现 1 次**（已 grep 实证）；环体、控制流、数据一字未改。
// 【为什么是这一枚 —— 未定价线索】本队 `1014` 台账把这一格列为**第一顺位未定价线索**：
//   签名平面构建环里的 8 条目标前瞻**距离是硬编码的 `+32`**（作者一次定死，从未在判题机扫过）。
//   在册同类证据：同族 `MF_INCA_PFD 16 → 32`（增量 OR 环 RMW 目标预取距离）判题机兑现
//   **−318.86 µs**（#129085 那一发）—— 说明本族「距离」旋钮在判题机上是**活的**；
//   而 N1 律指出「预取刀的有效距离必须 ≥ 该环单迭代拍数 × 访存延迟（拍）」。
//   本环每次迭代（j += 8）要做 **8 条随机 8 B 行 gather**（`cached_signature[od[j+e]*8]`，
//   e=0..7，8 个相异 cache line），环体约 60 条 uop ⇒ 若按 ~15~20 拍/迭代估，
//   距离 32（= 4 轮 ≈ 60~80 拍）**可能不足**一次 DRAM miss（~200 拍）⇒ 值得扫。
// 【改动性质】**纯预取**（只改预取地址，不改任何被读写的语义数据、不改控制流之外的任何东西）
//   ⇒ 输出**逐位必同**（值闸门 = 与基座 `.text` 的差异仅落在该 `disp32` 常量上 + 逐位哈希）。
// 【本发性质（锚件专用占位）】—（本件为**臂件**，非锚）
// 【目的（试验性提交须写）】把该前瞻距离在本题**现役底盘 + 生产规模（n=1e6）**上定价。
//   ★**现实预期（不声称能转绿）**：严支缺口 **31.705713 ms（0.674%）** 远大于单条预取距离的量级
//   ⇒ 本发的价值 = **压低 `mine` / 加宽武装窗**（被动窗下沿 = `(mine−1µs)/1.005`，
//   `mine` 每降 1 µs 窗即加宽 1 µs），**逐位等价 ⇒ 零下行风险**。
// 【预登记判据（先写死后发）】Δ 相对**同批同码锚带 S（双口径：批内极差 + 紧邻锚对）**：
//   `|Δ| ≤ 紧邻带` ⇒ 记「抖动，不认」；`Δ ≤ −3×紧邻带` ⇒ **必报（可判正）**；
//   任一枚 < 4701.587032 ms ⇒ **破纪录上报**；读数 ≤ **4669.881319 ms** ⇒ **翻色、立刻停手上报**。
// ================================================
// ===== REFERENCES =====
// [1] duck.ac 用户 saffah_cc_v41_agg1（本账号），提交 #128473
//     <https://duck.ac/submission/128473>（`1014` 现役最快件 / mine = 4701.976451 ms）。
//     —— **本件正文的基座**：代码正文为该提交的**逐字节副本**（仅其自带注释头被本段替换，
//     并加 `#line 397` 复原行号 ⇒ 第 397 行起逐字符不变）。本发相对基座**唯一改动 = 把 `#define MF_INCA_PFD 16` 改成 **32**（单常数，增量 OR 环 RMW 目标预取的前瞻距离；只影响预取地址，**逐位必同**）**。
// [2] **原始作者（引擎来源）**：duck.ac 用户 saffah_codex_6a_agg3，提交 #126143
//     <https://duck.ac/submission/126143>（`1014` 本题 T）。本件沿用的内容 = 该公开件的
//     全部引擎实现（位集平面 / 桶网格 / pass A 扫描 / fringe / 池构建族），经本账号公开提交
//     #126941 <https://duck.ac/submission/126941> → #127524 <https://duck.ac/submission/127524>
//     → #128321 <https://duck.ac/submission/128321> → #128473 逐代改良继承而来。
// [3] duck.ac 题目页 <https://duck.ac/problem/1014>（站点 FAQ 明示提交正文公开可见）。
// LICENSE：以上均为 duck.ac 公开提交，无第三方许可证声明、非来自任何外部仓库
//     ⇒ 按具名引用（账号 + 提交地址 + 沿用内容）使用，遵守其无附加限制的许可。
//     **本件未参考任何已发表论文 / 外部开源仓库。**
// ========== 思路 ========== 目的（试验性提交须写）
// 【本发单变量】把 `#define MF_INCA_PFD 16` 改成 **32**（单常数，增量 OR 环 RMW 目标预取的前瞻距离；只影响预取地址，**逐位必同**）。位点 = 基座第 450 行（本题最后一条 `#pragma GCC optimize`，
//   即 `("tracer")`）之后、`#include <type_traits>` 之前；`depth=0`、本件**无 `push_options`/
//   `pop_options`**（已 grep 实证 ⇒ pragma 全为文件级、对该点之后的全部函数生效）、不在任何
//   `#if` 内、目标函数全在其后 —— **独立新起一行**追加（律 N1：每旗标独立成行）。正文其余一字未改。
// 【为什么是这一枚】本队「2^k 阈值猎手」席 2026-10-16 的清单把这个旋钮列为**在册未定价**的一档：该环每次迭代约 8.3 拍（在册实测：L1 约 12 条 LFB ⇒ 同时约 6 次迭代在飞 × 约 40 拍 ≈ 8.3 拍/迭代 = 110 ms），距离 16 ≈ 133 拍 < DRAM 延迟约 200 拍 ⇒ 按同一算式 32 才够藏一次 DRAM miss；且源码注释与在册自述都写明：本族「距离类」参数**在判题机从未扫过**（在册三刀全用 16）。本发 = 单常数 16→32（逐位必同）。
// 【目的（试验性提交须写）】把该旗标在本题**现役底盘**上定价。★**现实预期**：本题严支缺口
//   **32.095132 ms（0.683%）** 远大于单枚旗标的量级 ⇒ 价值 = **压低 `mine` / 加宽武装窗**
//   （被动窗下沿 = (mine−1µs)/1.005，mine 每降 1 µs 窗口即加宽 1 µs）⇒ **本席不声称能转绿**。
// 【预登记判据（先写死后发）】Δ 相对**同批同码锚带 S**：|Δ| ≤ S ⇒ 记「抖动，不认」；
//   Δ ≤ −3·S ⇒ 必报；任一枚 < 4701.976451 ms ⇒ 破纪录上报；读数 ≤ **4669.881319 ms** ⇒ 翻色停手。
// ================================================
#line 397
#define MF_NOSDMP 1
//     `#108401` 之后新加入的四项**（见"思路"）；`#108502` 的 pass-A 取数环与我方
//     `#108401` 的 `[PFDENS]` 形态**逐字相同** ⇒ 对方已把我方 `[PFDENS]` 一并吸收。
// [3] 本账号 **saffah_cc_v41_agg1**，提交 **#108399 / #108359 / #108329**
//     <https://duck.ac/submission/108399> / <https://duck.ac/submission/108359> /
//     <https://duck.ac/submission/108329> —— 用途：本账号在本题的**上一代最好件链**
//     （对流内预取距离 `+48`、`MF_PFA 16→8`、以及对榜首 `#108184` 的复刻）。
// [4] `/home/yjp/duck.ac/problems/1015/notes.md`（本 lane 的全部判题机读数与已闭环轴）。
// ======================
// ===== 思路（历史席，正文中已并入）—— 详细论证见 problems/1014/notes.md 同名小节 =====
// ===== 思路（历史席，正文中已并入）—— 详细论证见 problems/1014/notes.md 同名小节 =====
// ===== 思路（历史席，正文中已并入）—— 详细论证见 problems/1014/notes.md 同名小节 =====
#pragma GCC optimize("O3,unroll-loops,no-strict-aliasing,schedule-insns,web")
#pragma GCC optimize("ira-loop-pressure")
#pragma GCC target("avx2,bmi,bmi2,popcnt,lzcnt,tune=skylake")
#define MF_QCOEF 140
#define MF_NOFILTB 1
// [x15b2/disp32] Compile-time signature-plane stride.  Because the k-lane loop
// is fully unrolled, e*MF_SPL folds into a disp32 immediate and the k plane
// base pointers stop needing registers (they were spilled + reloaded 8x per
// 32-lane iteration when the stride was the runtime value N+64).
// 1000064 == N + 64 for the judge configured n = 1e6, i.e. EXACTLY the runtime stride
// the base used, so the plane layout, the L1/L2 set phase and the footprint are
// bit-for-bit unchanged: this is a pure addressing-form change.
// Valid only while N + 64 <= MF_SPL -- the s_sig allocation refuses otherwise.
/* [B14AA-SPL4K] 1014b 转绿刀 [SPL4K] 的姊妹移植：`s_sig` 逐维签名平面的**行距必须取
4 KB 的整数倍**。原值 1000064 = 244*4096 + 640 ⇒ 第 e 张平面页内偏移 = 640*e mod 4096 =
{0,640,1280,1920,2560,3200,3840,384}，而 fringe 分类器**同时**读 7 张平面（同一
`sbase + e*MF_SPL + aj`，e=0..k-1、e!=d）⇒ 七条同相流挤在相近物理地址（mmmc4k「七条同相流
行距填充」律）。取 245*4096 = 1003520 ⇒ 七张平面基址全部 4 KB 对齐、物理地址翻过 12 位以上。
判题机 1014b 同形刀实测 −0.570 ms / −0.698%、板面兑现 0.98（#122788）。纯寻址，语义不动。 */
#ifndef MF_SPL_ALIGN
#define MF_SPL_ALIGN 1011712
#endif
#define MF_SPL ((size_t)MF_SPL_ALIGN)
// duck.ac 1012..1016 : k-D dominance counting (k=6..10).
//
// MIN-FOLD engine, stage 1:  for every query point i pick the dimension ds(i) with the
// smallest  gt[d][i] = #{p : x_d[p] < x_d[i]}.  Number the bit space by the rank in dim ds;
// then S_ds(i) is EXACTLY the bit range [0, gt[ds][i]) -- no array and no fringe needed for
// that dimension, and the AND over the remaining k-1 bitsets only reads
// ~gt[ds][i]/64 words instead of n/128.  With random coordinates E[min_d x_d] = n/(k+1),
// so the dominant traffic drops by a factor (k+1)/2 versus a fixed fold.
//
// Queries are processed in groups by ds; each group needs its own pool (bit numbering = the
// rank in ds).  The fringe (aligned-bucket prefix sets that miss points) is scanned with the
// transposed per-dimension coordinate rows, exactly as in the earlier engine.
#include <sys/mman.h>
#ifndef MADV_HUGEPAGE
#define MADV_HUGEPAGE 14
#endif
#include <immintrin.h>
#include <type_traits>
#pragma GCC optimize("tracer")
#include <type_traits>
#ifdef LOCAL_TEST
#include <cstdio>
#include <cstdlib>
#include <ctime>
#else
#ifdef __cplusplus
extern "C" void *malloc(unsigned long);
#else
void *malloc(unsigned long);
#endif
#endif
typedef unsigned u32;
typedef unsigned char u8;
typedef unsigned short u16;
typedef unsigned long long u64;
// g++-9 -O2 applies -mavx256-split-unaligned-load, so EVERY _mm256_loadu_si256
// compiles to "vmovdqu xmm; vinserti128" -- TWO loads plus a shuffle instead of one.
// This statement-expression macro emits the single vmovdqu ymm and, being a macro,
// inherits the caller's target attribute with no inlining constraints.
#define LDU256(p) ({ __m256i _v256; __asm__("vmovdqu %1, %0" : "=x"(_v256) : "m"(*(const __m256i *)(p))); _v256; })
#ifndef MAXN
#define MAXN 1000005
#endif
#define MAXK 8
#define STRIDE 8
// Bucket-count coefficient: q ~ sqrt(n) * MF_QCOEF/100.  Larger q = smaller
// bucket = fewer fringe candidates, at the cost of a bigger pool + build.
// Bucket count = sqrt(n) * MF_QCOEF/100 * kscale[k].  MEASURED OPTIMA (board):
//   n=1e6  k=6..8 : 122 default is fine (1013 7.491 s, 1014 8.503 s at 122)
//   n=1e6  k=9/10 : 92  (1016 10.888 s at 92; 10.959 at 122; 11.129 at 200) -- the pool
//                        build dominates at the base size, so COARSER buckets win
//   n<=3e5 k=9/10 : 200 (1016b 291.4 ms vs 326.6 at 122; 1015a 1.444, 1016a 1.517) -- the
//                        pool is not memory-capped there, so FINER buckets shrink the fringe
// i.e. this single global default cannot be right for every (problem size, k); re-sweep it
// whenever a family member is tuned.
#ifndef MF_APF
#define MF_APF 0
#endif
#ifndef MF_QCOEF
#define MF_QCOEF 140
#endif
// Transposed fringe rows: stride padded to 8 u32 (32 B) for k<=8 so every
// candidate load is one aligned, line-contained 32-byte vector load.
#ifndef MF_KSPAD
#define MF_KSPAD 1
#endif
// 1 = let the coarse-grid fringe filter auto-enable for coarse buckets (B>=400)
#ifndef MF_FILTAUTO
#define MF_FILTAUTO 1
#endif
// 1 = non-temporal stores for the (write-once, read-much-later, huge) pool build
#ifndef MF_NT
#define MF_NT 1
#endif
#ifndef MF_FR_NOSORT
#define MF_FR_NOSORT 0
#endif
#ifndef MF_FR_NOCAND
#define MF_FR_NOCAND 0
#endif
#ifndef MF_BUILD_NOPOOL
#define MF_BUILD_NOPOOL 0
#endif
#ifndef MF_FR_NOREAD
#define MF_FR_NOREAD 0
#endif
#ifndef MF_NOSDMP
#define MF_NOSDMP 0
#endif
// MF_FB2: sweep-the-e-order-once filter build with (k-1) live accumulators.
// MEASURED WORSE on the board (1016 12.580 -> 13.361 s, 1014 9.400 -> 10.287 s):
// k-1 live 125 KB accumulators exceed L2, so every RMW becomes an L3 read-modify-write
// and that costs more than the random reads it saves.  Kept only as a springboard.
#ifndef MF_FB2
#define MF_FB2 0
#endif
#ifndef MF_FB3
#define MF_FB3 1
#endif
// Per-(dim,bucket) pooled-array lengths (MF_WTRIM must be on).
#ifndef MF_PLEN
#define MF_PLEN 1
#endif
#ifndef MF_BUDGETMB
#define MF_BUDGETMB 1800
#endif
#ifndef MF_NOFILTB
#define MF_NOFILTB 0
#endif
// Lookahead distance for prefetching the scattered per-query metadata
// (s_pm record, s_gt threshold, s_bd bucket word) in pass B / pass A.
#ifndef MF_PFB
#define MF_PFB 6
// [PFSC] prefetch distance for the per-dim gt/bd block's s_cntv[] row fetch (0 = off)
#define MF_PFSC 32
// [PFSC2] prefetch distance for the scatter loop's s_tmp[] read-modify-write (0 = off)
#define MF_PFSC2 32
#endif
// Pass A: sort each ds group by the bucket of one non-fold dim and copy that
// dim's (dim,bucket) array prefix into a small cache-resident scratch so the
// ~137 queries sharing it do not each re-fetch ~70 KB from DRAM.  1 = on.
#ifndef MF_ASCR
#define MF_ASCR 1
#endif
// ============================================================================
// !! DO NOT ENABLE MF_WTRIM / MF_PLEN !!
// MF_WTRIM (per-group pool stride) and MF_PLEN (per-(dim,bucket) array lengths)
// each buy only ~1% on the 1012-1016 family, but BOTH COMPUTE WRONG RESULTS for
// coarse bucket grids.  Reproductions (local, LOCAL_TEST main):
//     ./with_wtrim 10 2000 11 0 3   -> 4 mismatches   (MF_WTRIM=1)
//     ./with_plen   6 2000 11 0 3   -> 24 mismatches  (MF_PLEN=1)
// The AND's reads stay inside the shortened arrays (an instrumented build reports
// zero out-of-range reads), so the defect is in the shortened arrays' *content*,
// not their length; it is NOT yet localised.  The fastest recorded board times
// (1012 6.753 s, 1016 12.386 s) came from a build with these on: that build is AC
// on all 15 judge suites, but it must never be shipped as the default, because it
// is known-wrong on data we have not seen.  Fixing this properly is worth ~1%.
// ============================================================================
// Trim the pool stride per ds group to the largest R actually read (-13% pool).
#ifndef MF_WTRIM
#define MF_WTRIM 0
#endif
#define AS_WORDS 20480
#ifndef MF_PFA
#define MF_PFA 2
#endif
#ifndef MF_FR_NOMETA
#define MF_FR_NOMETA 0
#endif
#ifndef MAXQ
#define MAXQ 4096
#endif
static u32 s_ord[MAXK][MAXN];
// [FRORD] per-dim array in dim-d order with the exact bucket packed into the high bits.
// Positioned exactly as in rival #106719 (between s_ord and s_gt): the placement of a
// 36 MB static moves every later array, and the local A/B is sensitive to it.
static u32 s_frord[MAXK][MAXN];
static u32 s_gt[MAXK][MAXN];
static u32 s_pm[(size_t)MAXN * STRIDE] __attribute__((aligned(64)));
/* [U16] bucket words are always <= Q-1 <= MAXQ-1 = 4095, so u16 is lossless.
Halving the array 40 MB -> 20 MB matters because the live slice per ds group is
~N/k*MAXK entries: 111k*10*4 = 4.4 MB (u32) vs 2.2 MB (u16) against a 6 MB L3. */
// [MIRROR] dim-major SoA slice for the RANDOM single-field readers of s_bd.
// s_bd (point-major, stride 20 B) is kept for the whole-record readers.
// BDR pads each dim row to a 64-B boundary so panels start line-aligned.
#define E6Z_BDR (((size_t)(MAXN) + 31u) & ~(size_t)31u)
static u16 s_bdT[(size_t)MAXK * E6Z_BDR];
static u16 s_bd[(size_t)MAXN * MAXK];
/* [b14ae_ SGC] 组号表 s_g 的值域只有 0..k（<=10）⇒ u32 的 4 MB 表缩成 **1 MB**：
   它在计数排序与 [FRORD] 构建环里都是**随机 gather**，4 MB + 8 MB 顺序/散布流 = 12 MB
   远超判题机 L3 6 MB ⇒ 表必被冲掉；缩到 1 MB 后表可常驻（b14ad_ 的「表 < L3」判据）。
   逐位等价：所有读点都是下标/比较，u8 提升为 int 后值完全一样。 */
static u8 s_g[MAXN] __attribute__((aligned(64)));
static u32 s_qord[MAXN], s_tmp[MAXN], s_cntv[MAXN];
static u16 s_bkt[MAXN];
static u16 s_bks[MAXN];   // [K3a] sort keys scattered into post-sort order
static u32 s_nid[MAXN];
static u32 s_pos[MAXK][MAXN];
static u32 s_posp[(size_t)MAXK * MAXN];   // s_posp[p*MAXK+d] = rank of p in dim d
#ifndef MF_INCA
#define MF_INCA 1          // [INCA-S] 定扎车道：平面落盘 -> 顺序流；pass A 拷贝 -> 增量 OR
#endif
#ifndef MF_FRPF
#define MF_FRPF 32     // FRORD 构建环的链式目标预取距离
#endif
#ifndef MF_INCA_PFD
#define MF_INCA_PFD 32      // 增量环 RMW 目标预取距离（a13ag_ 的定价配置 = 0）
#endif
static int g_inca = 0;     // [INCA-S] 运行期有效性（N <= AS_WORDS*64 且流已分配）
static u32 *s_ho = 0;      // [INCA-S] 定扎车道顺序流 s_ho[ds*N + j]（ds = 该群的折维）
static size_t s_ho_cap = 0;
static u64 s_facc[(size_t)(MAXK - 1) * (((MAXN) + 63) / 64 + 16)];
static u32 s_bnde[(size_t)MAXK * (MAXQ + 2)];   // per-dim coarse-cell boundary positions
static u64 *s_fbm = 0;
static u64 s_fbuf[8192] __attribute__((aligned(64)));
static __m256i khi8v, khi8hv;
static u32 g_Qc = 32;
static u32 g_fbm_W = 0;
static int g_use_filt = 0;
static int g_lexsort = 0;
static int g_nofilt = 0;
static u32 s_fpos[(MAXK + 1) * MAXQ];
static u32 s_bval[(MAXK + 1) * MAXQ];
static u64 *s_bs = 0;
static u64 s_acc[MAXN / 64 + 256] __attribute__((aligned(64)));
static u32 *s_sdmp[MAXK];
static u64 s_ascr[AS_WORDS] __attribute__((aligned(64)));
static u32 s_maxr[(size_t)MAXK * MAXK * MAXQ];
// ---- byte-signature coarse filter for the fringe scan -------------------
// The fringe scan is BYTE-bound, not instruction-bound: at k=10/n=1.5e5 it moves
// 6.6 GB at 19.9 GB/s (judge DRAM read roofline is 21.1 GB/s), and at k=9/n=2e5
// 8.6 GB at 18.4 GB/s.  Each candidate costs 4*k bytes of s_sdmp row.  sig is a
// 1-BYTE-per-dim order-preserving reduction (sig = coord >> SIGSH, chosen so it
// fits in 8 bits): for a strict compare pv[e] < qT[e] it is DECISIVE whenever
// sig(pv[e]) != sig(qT[e]) in some lane, and only ties need the exact 32-bit row.
// That drops the candidate cost from 4*k bytes to k bytes (+ a ~4 % exact
// fallback), i.e. ~4x less traffic on the dominant term.
static u8 *s_sig = 0;
static u8 b14af_csig[(size_t)MAXN*8] __attribute__((aligned(64)));
static int g_sig_on = 0;
static int g_sig_merged = 0;
static u32 g_sigsh = 0;
static u32 s_pofs[(size_t)MAXK * MAXQ];
static u32 s_plen[(size_t)MAXK * MAXQ];
static int g_plen_on = 0;
static u32 g_W, g_Q, g_B;
static u32 solve_generation;
static u32 g_KS;   // row stride of the transposed coordinate rows (64 B for k>8)
static int g_nosort = 0;
static int g_nosdmp = 0;
static int g_skip_and = 0, g_skip_fringe = 0;
#define FPOS(d, b) s_fpos[(size_t)(d) * MAXQ + (b)]
#define BVAL(d, b) s_bval[(size_t)(d) * MAXQ + (b)]
static void *pool_alloc(size_t bytes) {
void *raw = malloc(bytes + (2u << 20));
if (!raw) return 0;
void *p = (void *)(((size_t)raw + (2u << 20) - 1) & ~(size_t)((2u << 20) - 1));
{ void *m = p; size_t mb = bytes & ~(size_t)((2u << 20) - 1);
  if (mb) madvise(m, mb, MADV_HUGEPAGE); }
return p;
}
// Harley-Seal / Muła nibble-popcount.  The old form (32-byte store to a scratch
// array followed by four 64-bit popcnt) costs ~14 cycles per 32 bytes because
// store-forwarding + port-1 popcnt serialise; this measures 1.48x faster on a
// pure-L1 microbenchmark (work/c6d2/alucost.cpp: 19.27 -> 13.06 cyc per chunk).
// Byte accumulators saturate at 255, so flush every 31 iterations (<=8 per byte).
__attribute__((target("avx2,popcnt")))
static inline u32 hsum256(__m256i v) {
/* [A15E-HS] register-only reduction: the 32 B store + four 8 B reloads paid a
store-forwarding stall once per query.  Same sum mod 2^32 => bit-identical. */
__m128i lo = _mm256_castsi256_si128(v);
__m128i hi = _mm256_extracti128_si256(v, 1);
__m128i s = _mm_add_epi64(lo, hi);
return (u32)((u64)_mm_cvtsi128_si64(s) + (u64)_mm_extract_epi64(s, 1));
}
template<int KK>
__attribute__((target("avx2,popcnt")))
static u32 and_popcount_range(const u64 *const *bs, u32 Wr, u32 rem, const u64 *const *bsNext) {
const __m256i lm = _mm256_set1_epi8(0x0f);
const __m256i lk = _mm256_setr_epi8(0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4,
0,1,1,2,1,2,2,3,1,2,2,3,2,3,3,4);
const __m256i z = _mm256_setzero_si256();
/* [PFDENS] one iteration = 64 B (a whole line) per lane: the row coverage is
unchanged, but each 64 B line is prefetched ONCE instead of twice, and the lead
time is 2x (8 iterations' worth).  Two nibble-LUT accumulators keep the flush
cadence at 31 iterations (<=8 per byte each), so the sum is unchanged. */
__m256i tot = z, acc0 = z, acc1 = z;
u32 w = 0, cnt = 0, total = 0;
/* [A15F-HEAD] In the last 48 words of every stream the +48 prefetch lands PAST the
stream end, i.e. those slots are wasted.  Redirect exactly those slots (same
request count - a substitution, not an addition) to the NEXT query's stream head
lines 0..5, which its own in-loop prefetch can never cover (distance 48 => its
lines 0..5 are always demand misses).  bsNext is the already-staged pointer array,
so this costs no extra address arithmetic.  Prefetch only => bit-identical. */
u32 wlim = (Wr > 96u) ? (Wr - 96u) : 0u;
for (; w + 8 <= wlim; w += 8) {
if (KK > 1) _mm_prefetch((const char *)(bs[1] + w + 96), _MM_HINT_T0);
if (KK > 2) _mm_prefetch((const char *)(bs[2] + w + 96), _MM_HINT_T0);
if (KK > 3) _mm_prefetch((const char *)(bs[3] + w + 96), _MM_HINT_T0);
if (KK > 4) _mm_prefetch((const char *)(bs[4] + w + 96), _MM_HINT_T0);
if (KK > 5) _mm_prefetch((const char *)(bs[5] + w + 96), _MM_HINT_T0);
if (KK > 6) _mm_prefetch((const char *)(bs[6] + w + 96), _MM_HINT_T0);
if (KK > 7) _mm_prefetch((const char *)(bs[7] + w + 96), _MM_HINT_T0);
__m256i a = LDU256((bs[0] + w));
for (int d = 1; d < KK; d++) a = _mm256_and_si256(a, LDU256((bs[d] + w)));
__m256i lo0 = _mm256_and_si256(a, lm);
__m256i hi0 = _mm256_and_si256(_mm256_srli_epi16(a, 4), lm);
acc0 = _mm256_add_epi8(acc0, _mm256_add_epi8(_mm256_shuffle_epi8(lk, lo0), _mm256_shuffle_epi8(lk, hi0)));
__m256i b = LDU256((bs[0] + w + 4));
for (int d = 1; d < KK; d++) b = _mm256_and_si256(b, LDU256((bs[d] + w + 4)));
__m256i lo1 = _mm256_and_si256(b, lm);
__m256i hi1 = _mm256_and_si256(_mm256_srli_epi16(b, 4), lm);
acc1 = _mm256_add_epi8(acc1, _mm256_add_epi8(_mm256_shuffle_epi8(lk, lo1), _mm256_shuffle_epi8(lk, hi1)));
if (++cnt == 31) {
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc0, z));
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc1, z));
acc0 = acc1 = z; cnt = 0;
}
}
if (bsNext) {
u32 off = 0;
for (; w + 8 <= Wr; w += 8, off += 8) {
if (KK > 1) _mm_prefetch((const char *)(bsNext[1] + off), _MM_HINT_T0);
if (KK > 2) _mm_prefetch((const char *)(bsNext[2] + off), _MM_HINT_T0);
if (KK > 3) _mm_prefetch((const char *)(bsNext[3] + off), _MM_HINT_T0);
if (KK > 4) _mm_prefetch((const char *)(bsNext[4] + off), _MM_HINT_T0);
if (KK > 5) _mm_prefetch((const char *)(bsNext[5] + off), _MM_HINT_T0);
if (KK > 6) _mm_prefetch((const char *)(bsNext[6] + off), _MM_HINT_T0);
if (KK > 7) _mm_prefetch((const char *)(bsNext[7] + off), _MM_HINT_T0);
__m256i a = LDU256((bs[0] + w));
for (int d = 1; d < KK; d++) a = _mm256_and_si256(a, LDU256((bs[d] + w)));
__m256i lo0 = _mm256_and_si256(a, lm);
__m256i hi0 = _mm256_and_si256(_mm256_srli_epi16(a, 4), lm);
acc0 = _mm256_add_epi8(acc0, _mm256_add_epi8(_mm256_shuffle_epi8(lk, lo0), _mm256_shuffle_epi8(lk, hi0)));
__m256i b = LDU256((bs[0] + w + 4));
for (int d = 1; d < KK; d++) b = _mm256_and_si256(b, LDU256((bs[d] + w + 4)));
__m256i lo1 = _mm256_and_si256(b, lm);
__m256i hi1 = _mm256_and_si256(_mm256_srli_epi16(b, 4), lm);
acc1 = _mm256_add_epi8(acc1, _mm256_add_epi8(_mm256_shuffle_epi8(lk, lo1), _mm256_shuffle_epi8(lk, hi1)));
if (++cnt == 31) {
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc0, z));
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc1, z));
acc0 = acc1 = z; cnt = 0;
}
}
}
for (; w + 8 <= Wr; w += 8) {
__m256i a = LDU256((bs[0] + w));
for (int d = 1; d < KK; d++) a = _mm256_and_si256(a, LDU256((bs[d] + w)));
__m256i lo0 = _mm256_and_si256(a, lm);
__m256i hi0 = _mm256_and_si256(_mm256_srli_epi16(a, 4), lm);
acc0 = _mm256_add_epi8(acc0, _mm256_add_epi8(_mm256_shuffle_epi8(lk, lo0), _mm256_shuffle_epi8(lk, hi0)));
__m256i b = LDU256((bs[0] + w + 4));
for (int d = 1; d < KK; d++) b = _mm256_and_si256(b, LDU256((bs[d] + w + 4)));
__m256i lo1 = _mm256_and_si256(b, lm);
__m256i hi1 = _mm256_and_si256(_mm256_srli_epi16(b, 4), lm);
acc1 = _mm256_add_epi8(acc1, _mm256_add_epi8(_mm256_shuffle_epi8(lk, lo1), _mm256_shuffle_epi8(lk, hi1)));
if (++cnt == 31) {
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc0, z));
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc1, z));
acc0 = acc1 = z; cnt = 0;
}
}
for (; w + 4 <= Wr; w += 4) {
__m256i a = LDU256((bs[0] + w));
for (int d = 1; d < KK; d++) a = _mm256_and_si256(a, LDU256((bs[d] + w)));
__m256i lo2 = _mm256_and_si256(a, lm);
__m256i hi2 = _mm256_and_si256(_mm256_srli_epi16(a, 4), lm);
acc0 = _mm256_add_epi8(acc0, _mm256_add_epi8(_mm256_shuffle_epi8(lk, lo2), _mm256_shuffle_epi8(lk, hi2)));
}
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc0, z));
tot = _mm256_add_epi64(tot, _mm256_sad_epu8(acc1, z));
total = hsum256(tot);
for (; w < Wr; w++) {
u64 v = bs[0][w];
for (int d = 1; d < KK; d++) v &= bs[d][w];
total += (u32)__builtin_popcountll(v);
}
if (rem) {
u64 v = bs[0][Wr];
for (int d = 1; d < KK; d++) v &= bs[d][Wr];
total += (u32)__builtin_popcountll(v & ((1ull << rem) - 1));
}
return total;
}
static const u32 g_r2j_kmaskT[32]={0u,1u,3u,7u,15u,31u,63u,127u,255u,511u,1023u,2047u,
4095u,8191u,16383u,32767u,65535u,131071u,262143u,524287u,1048575u,2097151u,
4194303u,8388607u,16777215u,33554431u,67108863u,134217727u,268435455u,
536870911u,1073741823u,2147483647u};
#define TIE_CAP 524288u
#define TIE_EW 12
alignas(64) static u32 tie_data[(size_t)TIE_CAP*TIE_EW],tie_points[TIE_CAP];
static u32 tie_count;
static void flush_ties(u32 *out){
 u32 n=tie_count;tie_count=0;
 const __m256i all8=_mm256_set1_epi32(-1);
 for(u32 b=0;b<n;b+=256){
  u32 end=b+256<n?b+256:n;
  for(u32 j=b;j<end;j++){
   const u32 *e=tie_data+(size_t)j*TIE_EW;
   u32 p=s_ord[e[2]][e[1]];tie_points[j]=p;
   _mm_prefetch((const char*)(s_pm+(size_t)p*STRIDE),_MM_HINT_T0);
  }
  for(u32 j=b;j<end;j++){
   const u32 *e=tie_data+(size_t)j*TIE_EW;
   __m256i v=LDU256(s_pm+(size_t)tie_points[j]*STRIDE);
   __m256i q=_mm256_loadu_si256((const __m256i*)(e+3));
   out[e[0]]+=(u32)_mm256_testc_si256(_mm256_cmpgt_epi32(q,v),all8);
  }
 }
}
template<int KK>
__attribute__((target("avx2,popcnt")))
static void query_group(u32 N, u32 *out, u32 ds, u32 q0, u32 q1, u32 Q, u32 W) {
const int k = KK;
const int K1 = KK - 1;
// [AV8] avpool row stride in u32.  For KK <= 9 exmask never reaches bit 8, so
// avpool lanes 8..15 are dead (proof at the qTh site) and half a row suffices.
// KK == 10 can reach bit 8, so it keeps the original 16-wide row.
const u32 ASTRL = (KK <= 9) ? 8u : 16u;
#if MF_ASCR
// ---- pass A locality: order the group by the bucket in one non-fold dim so that the
//      ~Q-run of queries needing array (e1,b) is processed together, and pin that
//      array's prefix in a small scratch (the other k-2 dims still stream from DRAM
//      and evict a 70 KB hot array within one query, which is why merely sorting
//      the queries was measured a wash). ----
const u32 d1 = (ds == 0) ? 1u : 0u;
const u32 e1 = (d1 < ds) ? d1 : d1 - 1;
{
/* [B14V-FRSORT] s_frord[d1] already holds group ds in dim-d1 bucket order with the
bucket packed in the high bits (built once), so the per-group counting sort and
its random s_bdT key reads are redundant. */
const u32 *const fo1 = s_frord[d1];
for (u32 qi = q0; qi < q1; qi++) s_qord[qi] = fo1[qi] & 0xFFFFFu;
}
#else
const u32 d1 = 0, e1 = 0;
#endif
// ---- pass A: aligned AND; queries ordered lexicographically by their bucket vector so
//      that consecutive queries share bitset arrays (keeps the working set small). ----
if (g_lexsort) {
u32 *cntv_ = s_cntv;
for (int dd = k - 1; dd >= 0; dd--) {
if (dd == ds) continue;
for (u32 v = 0; v <= Q; v++) cntv_[v] = 0;
for (u32 qi = q0; qi < q1; qi++) cntv_[s_bdT[(size_t)dd * E6Z_BDR + s_qord[qi]] + 1]++;
{ u32 ac = q0; for (u32 v = 0; v <= Q; v++) { u32 c = cntv_[v]; cntv_[v] = ac; ac += c; } }
for (u32 qi = q0; qi < q1; qi++) { u32 id = s_qord[qi]; s_tmp[cntv_[s_bdT[(size_t)dd * E6Z_BDR + id] + 1]++] = id; }
for (u32 qi = q0; qi < q1; qi++) s_qord[qi] = s_tmp[qi];
}
}
#ifdef PF_DIST
for (u32 qi = q0; qi < q0 + PF_DIST && qi + PF_DIST < q1; qi++) {
u32 j = s_qord[qi + PF_DIST];
const u16 *bdn = s_bd + (size_t)j * MAXK;
for (int e = 0; e < K1; e++) {
int d = (e < ds) ? e : e + 1;
const char *pp = (const char *)(s_bs + ((size_t)e * (Q + 1) + bdn[d]) * W);
_mm_prefetch(pp, _MM_HINT_T0);
_mm_prefetch(pp + 64, _MM_HINT_T0);
_mm_prefetch(pp + 128, _MM_HINT_T0);
}
}
#endif
#if MF_ASCR
{
#if MF_INCA
/* [INCA-S] 增量维护 s_ascr：每群零化一次，随后按 b1 升序把 delta = 点集
{ j ∈ [pend(prev), pend(b1)) }（dim-d1 序的连续段）顺序 OR 进去 —— 取数走池构建
写好的顺序流 s_ho[ds*N + j]。scratch 覆盖整个位空间（AS_WORDS*64 = 1,310,720 ≥ N）
⇒ 任意 R 的 [0,R) 都与原平面逐位相同。 */
const u32 *a13_ho = 0;
u32 a13_app = 0;
if (g_inca) {
a13_ho = s_ho;
for (u32 w = 0; w < AS_WORDS; w++) s_ascr[w] = 0;
}
#endif
u32 qi = q0;
while (qi < q1) {
const u32 *const fo1b = s_frord[d1];
u32 b1 = fo1b[qi] >> 20;
u32 qe = qi + 1;
while (qe < q1 && (fo1b[qe] >> 20) == b1) qe++;
const u64 *bp1 = g_plen_on ? (s_bs + s_pofs[(size_t)e1 * MAXQ + b1])
: (s_bs + ((size_t)e1 * (Q + 1) + b1) * W);
#if MF_INCA
if (a13_ho) {
u32 pto = (b1 < Q) ? FPOS(d1, b1) : (u32)N;
const u32 *hp = a13_ho;
#if MF_INCA_PFD > 0
while (a13_app < pto) {
if (a13_app + MF_INCA_PFD < pto)
_mm_prefetch((const char *)(s_ascr + (hp[a13_app + MF_INCA_PFD] >> 6)), _MM_HINT_T0);
u32 q = hp[a13_app]; s_ascr[q >> 6] |= 1ull << (q & 63); a13_app++; }
#else
while (a13_app < pto) { u32 q = hp[a13_app]; s_ascr[q >> 6] |= 1ull << (q & 63); a13_app++; }
#endif
bp1 = s_ascr;
} else
#endif
{
u32 mx = 1;
for (u32 t = qi; t < qe; t++) { u32 rr = s_gt[ds][s_qord[t]]; if (rr > mx) mx = rr; }
u32 nwd = (mx >> 6) + 1;
if (nwd <= AS_WORDS) {
const u64 *src = bp1;
for (u32 w = 0; w < nwd; w++) s_ascr[w] = src[w];
bp1 = s_ascr;
}
}
/* [A15D-PIPE] entry-chain software pipeline: the CURRENT query's plane
pointers are computed one iteration EARLIER (staged into the other buffer),
so the serial chain s_qord->s_bd->s_pofs->bp overlaps the previous query's
scan instead of preceding its own.  Bit-identical (same pointers). */
const u64 *bp2v[2][MAXK];
u32 Rcur = s_gt[ds][s_qord[qi]];
{
const u16 *bd0 = s_bd + (size_t)s_qord[qi] * MAXK;
for (int e = 0; e < K1; e++) {
if (e == (int)e1) bp2v[qi & 1u][e] = bp1;
else { int d = (e < ds) ? e : e + 1;
bp2v[qi & 1u][e] = g_plen_on ? (s_bs + s_pofs[(size_t)e * MAXQ + bd0[d]])
: (s_bs + ((size_t)e * (Q + 1) + bd0[d]) * W); }
}
}
for (u32 qq = qi; qq < qe; qq++) {
u32 i = s_qord[qq];
#if MF_PFA > 0
{
u32 qj = qq + MF_PFA;
if (qj < qe) {
u32 j = s_qord[qj];
_mm_prefetch((const char *)(s_bd + (size_t)j * MAXK), _MM_HINT_T0);
_mm_prefetch((const char *)(s_gt[ds] + j), _MM_HINT_T0);
// [PFSC-B] pass-B's non-chained random RMW target out[j] was missing from
// this block (sibling 1015b measured the same omission at -1.486%).  j comes
// from the SEQUENTIAL array s_qord, so the lead-MF_PFA line needs no
// second-level dependency (BRIEF 2.19.269 chained-cost law / 2.19.323).
__asm__ __volatile__("prefetchw %0" :: "m"(*(const char *)(out + j)));  /* [A15E-PW] pass-A store target: exclusive */
}
}
#endif
if (qq + 1 < qe) {
const u16 *bdn = s_bd + (size_t)s_qord[qq + 1] * MAXK;
const u64 **bn = bp2v[(qq + 1) & 1u];
for (int e = 0; e < K1; e++) {
if (e == (int)e1) bn[e] = bp1;
else { int d = (e < ds) ? e : e + 1;
bn[e] = g_plen_on ? (s_bs + s_pofs[(size_t)e * MAXQ + bdn[d]])
: (s_bs + ((size_t)e * (Q + 1) + bdn[d]) * W); }
}
}
/* [A15D-PIPE2] also stage the gating rank: Rnext is loaded during the
previous iteration, so the current query's critical path is only
(scan -> out[i]) with no leading random load.  Bit-identical. */
const u32 R = Rcur;
if (qq + 1 < qe) Rcur = s_gt[ds][s_qord[qq + 1]];
u32 cnt = 0;
if (R && !g_skip_and) {
cnt = and_popcount_range<KK - 1>(bp2v[qq & 1u], R >> 6, R & 63u,
(qq + 1 < qe) ? bp2v[(qq + 1) & 1u] : 0);
}
out[i] = cnt;
}
qi = qe;
}
}
#else
for (u32 qi = q0; qi < q1; qi++) {
u32 i = s_qord[qi];
const u16 *bd = s_bd + (size_t)i * MAXK;
#if MF_PFA > 0
{
u32 qj = qi + MF_PFA;
if (qj < q1) {
u32 j = s_qord[qj];
_mm_prefetch((const char *)(s_bd + (size_t)j * MAXK), _MM_HINT_T0);
_mm_prefetch((const char *)(s_gt[ds] + j), _MM_HINT_T0);
// [PFSC-B] pass-B's non-chained random RMW target out[j] was missing from
// this block (sibling 1015b measured the same omission at -1.486%).  j comes
// from the SEQUENTIAL array s_qord, so the lead-MF_PFA line needs no
// second-level dependency (BRIEF 2.19.269 chained-cost law / 2.19.323).
_mm_prefetch((const char *)(out + j), _MM_HINT_T0);
}
}
#endif
#ifdef PF_DIST
if (qi + PF_DIST < q1) {
u32 j = s_qord[qi + PF_DIST];
const u16 *bdn = s_bd + (size_t)j * MAXK;
for (int e = 0; e < K1; e++) {
int d = (e < ds) ? e : e + 1;
const char *pp = (const char *)(s_bs + ((size_t)e * (Q + 1) + bdn[d]) * W);
_mm_prefetch(pp, _MM_HINT_T0);
_mm_prefetch(pp + 64, _MM_HINT_T0);
_mm_prefetch(pp + 128, _MM_HINT_T0);
}
}
#endif
const u32 R = s_gt[ds][i];
u32 cnt = 0;
if (R && !g_skip_and) {
const u64 *bp[MAXK];
for (int e = 0; e < K1; e++) {
int d = (e < ds) ? e : e + 1;
bp[e] = g_plen_on ? (s_bs + s_pofs[(size_t)e * MAXQ + bd[d]])
: (s_bs + ((size_t)e * (Q + 1) + bd[d]) * W);
#ifdef DBG_PLEN
if (g_plen_on) {
u32 need = (R >> 6) + ((R & 63) ? 1 : 0);
if (need > s_plen[(size_t)e * MAXQ + bd[d]])
printf("DBG ds=%d e=%d d=%d b=%u R=%u need=%u len=%u\n", ds, e, d, bd[d], R, need, s_plen[(size_t)e * MAXQ + bd[d]]);
}
#endif
}
cnt = and_popcount_range<KK - 1>(bp, R >> 6, R & 63u, 0);
}
out[i] = cnt;
}
#endif
if (g_skip_fringe) return;
static u32 *avpool = 0; static size_t avcap = 0;
{
size_t needav = (size_t)N * STRIDE * 4 + 256;
static u32 avN = 0;
if (needav > avcap) { avpool = (u32 *)pool_alloc(needav); avcap = avpool ? needav : 0; avN = 0; }
/* [b14z_ AVONCE] `av[ds]` 是死值：唯一消费者 `Avm = LDU256(Avp) | notexlo`，而
`notexlo[ds] = 0x7FFFFFFF` 由构造保证（`exmask` 只置 `e < d 且 e != ds` 的位
⇒ `hiA[ds] = 0x7FFFFFFF`）⇒ 只要 `av[ds] < 0x80000000`（BVAL 是坐标 < N，或 0），
`Avm[ds]` 恒 0x7FFFFFFF ⇒ **整行与 ds 无关** ⇒ 每个 N 只建一次（8 -> 1），
且写入由 s_qord 随机序 32 B 散写变为 ii = 0..N-1 顺序写。 */
if (avpool && avN != solve_generation) {
for (u32 ii = 0; ii < N; ii++) {
const u16 *bdi = s_bd + (size_t)ii * MAXK;
u32 *av = avpool + (size_t)ii * ASTRL;
for (int e = 0; e < k && e < (int)ASTRL; e++) av[e] = BVAL(e, bdi[e]);
for (int e = k; e < (int)ASTRL; e++) av[e] = 0x7FFFFFFFu;
}
avN = solve_generation;
}
}
// ---- pass B: fringe, dimension-major.  Ordering the group's queries by their bucket in
//      dim d makes every scan walk one L1-resident bucket block of the dim-d order. ----
auto fringe_dim = [&](auto dim) {
constexpr int d=decltype(dim)::value;
if(d==ds) return;
// [FRORD] the per-dim counting sort is GONE: s_frord[d] is already in dim-d order
// with each point's bucket packed in the high bits, so the loop below reads it
// sequentially instead of gather-sorting s_qord through random s_bdT buckets.
const u32 *fo = s_frord[d];
u32 exmask = 0;
for (int e = 0; e < d; e++) if (e != ds) exmask |= 1u << e;
// "don't care" lanes are forced to a maximum-unsigned value, so one unsigned max test
// per candidate decides domination *and* the exclusion rule together.
u32 hiA[8], hiAh[8];
for (int e = 0; e < 8; e++) hiA[e] = ((exmask >> e) & 1u) ? 0u : 0x7FFFFFFFu;
for (int e = 8; e < 16; e++) hiAh[e - 8] = ((exmask >> e) & 1u) ? 0u : 0x7FFFFFFFu;
const __m256i notexlo = LDU256(hiA);
const __m256i notexloh = LDU256(hiAh);
const __m256i all8 = _mm256_set1_epi32(-1);
// [x15b2/disp32] one base register; the per-plane offset e*MF_SPL is a
// compile-time constant and therefore a disp32 immediate in the load.
const u8 *const sbase = s_sig + (size_t)d * (k-1) * MF_SPL;
const u32 *const odd = s_ord[d];
for (u32 qi = q0; qi < q1; qi++) {
u32 i = fo[qi] & 0xFFFFFu;
#if MF_PFB > 0
{
u32 qj = qi + MF_PFB;
if (qj < q1) {
u32 j = fo[qj] & 0xFFFFFu;
_mm_prefetch((const char *)(s_pm + (size_t)j * STRIDE), _MM_HINT_T0);
_mm_prefetch((const char *)(s_gt[d] + j), _MM_HINT_T0);
_mm_prefetch((const char *)(avpool + (size_t)j * ASTRL), _MM_HINT_T0);
/* [PFB3-OUT / 2.19.339+2.19.344] pass-B's RMW target out[] was absent here: the
body does out[i] += add (lines 779/950) and the block prefetched only the
metadata rows.  RMW + 1.2 MB (n=3e5) + lookahead MF_PFB steps.  Same cell that
paid -24.1 ms (#106747) / -19.4 ms (#106762, green) on 1015a; the pass-A
store variant on this problem was only -4.5 ms.  Pure prefetch, bit-identical. */
__asm__ __volatile__("prefetchw %0" :: "m"(*(const char *)(out + j)));
}
}
#endif
const u32 lo = FPOS(d, fo[qi] >> 20);   // [K3a] sequential key, not a cold s_bd line
const u32 hi = s_gt[d][i];
if (lo >= hi) continue;
const u32 *rec = s_pm + (size_t)i * STRIDE;
const u32 *Avp = avpool + (size_t)i * ASTRL;
__m256i qr = _mm256_or_si256(_mm256_load_si256((const __m256i *)rec), khi8v);
__m256i Avm = _mm256_or_si256(LDU256(Avp), notexlo);
__m256i qT = _mm256_min_epi32(qr, Avm);
__m256i qTh;
if (k > 9) {
// k = 10: d can reach 9, so exmask bit 8 is reachable -> keep the real test.
__m256i qrh = _mm256_or_si256(LDU256((rec + 8)), khi8hv);
__m256i Avmh = _mm256_or_si256(LDU256((Avp + 8)), notexloh);
qTh = _mm256_min_epi32(qrh, Avmh);
} else if (k > 8) {
// [AV8] k = 9: expmask stays inside bits 0..7 => hiAh is all-0x7FFFFFFF
// => Avmh is identically 0x7FFFFFFF => qTh = min_epi32(qrh, 0x7FFFFFFF)
// == qrh.  avpool lanes 8..15 (and this load + min) are dead code.
qTh = _mm256_or_si256(LDU256((rec + 8)), khi8hv);
} else {
qTh = all8;
}
u32 add = 0;
const u32 *c0 = s_sdmp[d] ? (s_sdmp[d] + (size_t)lo * g_KS) : 0;
if (g_use_filt && KK == 9) {
const u32 w0 = lo >> 6;
const u32 nw = ((hi - 1) >> 6) - w0 + 1;
const u64 wcs = (u64)N / g_Qc + 1;
const u32 FW = g_fbm_W;
const u64 *rows[8];
int nr = 0;
for (int e = 0; e < 9; e++) {
if (e == d) continue;
u32 cell = (u32)((u64)rec[e] / wcs) + 1;
if (cell > g_Qc) cell = g_Qc;
rows[nr++] = s_fbm + ((size_t)(d * 9 + e) * (g_Qc + 1) + cell) * FW + w0;
}
const u64 *r0 = rows[0];
const u64 *r1 = rows[1];
const u64 *r2 = rows[2];
const u64 *r3 = rows[3];
const u64 *r4 = rows[4];
const u64 *r5 = rows[5];
const u64 *r6 = rows[6];
const u64 *r7 = rows[7];
const u32 remh = hi & 63u;
{
u64 vv = r0[0] & r1[0] & r2[0] & r3[0] & r4[0] & r5[0] & r6[0] & r7[0];
vv &= ~0ull << (lo & 63);
if (nw == 1u && remh) vv &= (1ull << remh) - 1;
u32 word = w0;
#if MF_FR_NOCAND
vv = 0;
#endif
while (vv) {
int t = __builtin_ctzll(vv);
vv &= vv - 1;
u32 j = (word << 6) + (u32)t;
const u32 *cp = c0 ? (c0 + (size_t)(j - lo) * g_KS) : (s_pm + (size_t)s_ord[d][j] * STRIDE);
__m256i pv = LDU256(cp);
__m256i pvh = LDU256(cp + 8);
u32 ok = (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
ok &= (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
add += ok;
}
}
for (u32 w = 1; w + 1 < nw; w++) {
u64 vv = r0[w] & r1[w] & r2[w] & r3[w] & r4[w] & r5[w] & r6[w] & r7[w];
u32 word = w0 + w;
#if MF_FR_NOCAND
vv = 0;
#endif
while (vv) {
int t = __builtin_ctzll(vv);
vv &= vv - 1;
u32 j = (word << 6) + (u32)t;
const u32 *cp = c0 ? (c0 + (size_t)(j - lo) * g_KS) : (s_pm + (size_t)s_ord[d][j] * STRIDE);
__m256i pv = LDU256(cp);
__m256i pvh = LDU256(cp + 8);
u32 ok = (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
ok &= (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
add += ok;
}
}
if (nw > 1u) {
u32 w = nw - 1u;
u64 vv = r0[w] & r1[w] & r2[w] & r3[w] & r4[w] & r5[w] & r6[w] & r7[w];
if (remh) vv &= (1ull << remh) - 1;
u32 word = w0 + w;
#if MF_FR_NOCAND
vv = 0;
#endif
while (vv) {
int t = __builtin_ctzll(vv);
vv &= vv - 1;
u32 j = (word << 6) + (u32)t;
const u32 *cp = c0 ? (c0 + (size_t)(j - lo) * g_KS) : (s_pm + (size_t)s_ord[d][j] * STRIDE);
__m256i pv = LDU256(cp);
__m256i pvh = LDU256(cp + 8);
u32 ok = (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
ok &= (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
add += ok;
}
}
out[i] += add;
continue;
}
if (g_use_filt) {
const u32 w0 = lo >> 6;
const u32 nw = ((hi - 1) >> 6) - w0 + 1;
u64 *fb = s_fbuf;
const u64 wcs = (u64)N / g_Qc + 1;
const u32 FW = g_fbm_W;
int first = 1;
#if MF_FR_NOREAD
first = 0;
for (u32 w = 0; w < nw; w++) fb[w] = 0;
#endif
for (int e = 0; e < k; e++) {
if (e == d) continue;
u32 c = (u32)((u64)rec[e] / wcs) + 1;
if (c > g_Qc) c = g_Qc;
const u64 *row = s_fbm + ((size_t)(d * k + e) * (g_Qc + 1) + c) * FW + w0;
if (first) { for (u32 w = 0; w < nw; w++) fb[w] = row[w]; first = 0; }
else { for (u32 w = 0; w < nw; w++) fb[w] &= row[w]; }
}
fb[0] &= ~0ull << (lo & 63);
u32 r2 = hi & 63u;
if (r2) fb[nw - 1] &= (1ull << r2) - 1;
for (u32 w = 0; w < nw; w++) {
u64 vv = fb[w];
#if MF_FR_NOCAND
vv = 0;
#endif
while (vv) {
int t = __builtin_ctzll(vv);
vv &= vv - 1;
u32 j = ((w0 + w) << 6) + (u32)t;
const u32 *cp = c0 ? (c0 + (size_t)(j - lo) * g_KS) : (s_pm + (size_t)s_ord[d][j] * STRIDE);
__m256i pv = LDU256(cp);
u32 ok = (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
if (k > 8) {
__m256i pvh = LDU256((cp + 8));
ok &= (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
}
add += ok;
}
}
out[i] += add;
continue;
}
if (g_sig_on && (hi - lo) >= 2) {
// ---- coarse byte-signature scan, exact fallback only on ties ----
// (no qtv store: the broadcast signatures are built in the vector domain)
__m256i qsb[16];
// Lane d needs NO test at all: the band is [FPOS(d,bd_d), gt_d(i)) in dim-d
// order, and gt_d(i) is the START of i's tie group, so every candidate in the
// band has x_d[c] < x_d[i] = qT[d] by construction.  Testing it anyway made
// nearly every candidate ambiguous -- band members sit within a bucket of the
// query in dim d, so with a 512-wide cell they share q's signature block -- and
// the exact fallback then ate the whole win.  Neutralising the lane is EXACT.
{
const __m256i zero8 = _mm256_setzero_si256();
const __m256i shv = _mm256_set1_epi32((int)g_sigsh);
// [f5/setup-hoist] sv depends on the half (e<8) only, not on e itself.
/* [ONEACC] qsb is built from sig-1 so that one OR chain can carry the
three states (0 = strictly less, 1 = tie, >=2 = strictly greater). */
const __m256i bias8 = _mm256_set1_epi8((char)0x80);
const __m256i sv_lo = _mm256_srlv_epi32(qT, shv);
const __m128i u16=_mm_packus_epi32(_mm256_castsi256_si128(sv_lo),_mm256_extracti128_si256(sv_lo,1));
const __m128i u8s=_mm_packus_epi16(u16,u16);
const __m256i packed=_mm256_xor_si256(_mm256_broadcastsi128_si256(u8s),bias8);
for (int e = 0; e < k; e++) {
if(e==(int)d)continue;
qsb[e-(e>(int)d)]=_mm256_shuffle_epi8(packed,_mm256_set1_epi8((char)e));
}
}
// [f5/nocmp] lane d is provably true for every candidate of this band (the
// [nolane] argument), so the loop below simply skips it -- the branch is
// data-independent and perfectly predicted.  The former pre-pass that copied
// the k-1 live planes and their broadcasts into two stack arrays is gone: it
// cost 2*(k-1) stores plus the reload of every ymm per (query,dim), and the
// measured cost of that form on the sibling rows is large (1015a +2.42%,
// 1015b +4.30% for ADDING it -- see #103367 / #103361).
u32 aj = lo, aacc = 0, aexact = 0;
for (; aj + 32 <= hi; aj += 32) {

/* [ONEACC] one accumulator.  Byte 0 / 1 / >=2 = strictly-less / tie /
strictly-greater, and an OR cannot turn one of those into another.
decm == the old (ltm & ~amb) and amb == the old amb, bit for bit. */
__m256i ac=_mm256_subs_epi8(LDU256(sbase+aj),qsb[0]);
for(int e=1;e<k-1;e++){
__m256i cv=LDU256(sbase+(size_t)e*MF_SPL+aj);
ac=_mm256_max_epi8(ac,_mm256_subs_epi8(cv,qsb[e]));
}
u32 decm = (u32)_mm256_movemask_epi8(ac);
u32 amb = (u32)_mm256_movemask_epi8(_mm256_cmpeq_epi8(ac, _mm256_setzero_si256()));
aacc += (u32)__builtin_popcount(decm);
while (amb) {
int t = __builtin_ctz(amb); amb &= amb - 1;
if(KK==8 && !c0 && tie_count<TIE_CAP){
 u32 *e=tie_data+(size_t)tie_count++*TIE_EW;
 e[0]=i;e[1]=aj+(u32)t;e[2]=d;
 _mm256_storeu_si256((__m256i*)(e+3),qT);
 continue;
}
const u32 *cp = c0 ? (c0 + (size_t)(aj + (u32)t - lo) * g_KS) : (s_pm + (size_t)odd[aj + (u32)t] * STRIDE);
__m256i pv = LDU256(cp);
u32 okl = (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
if (k > 8) {
__m256i pvh = LDU256((cp + 8));
okl &= (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
}
aacc += okl;
}
}
// ---- tail: ONE overlapping 32-lane chunk, masked.  Instrumented counts
//      showed (hi-lo) mod 32 = 10.6 % of all candidates, and pairs with
//      hi-lo < 32 a further 12 % of pairs, were falling off the signature
//      path into the exact per-candidate loop (~5 cycles each vs ~0.6 for
//      the vector path).  Re-reading the last 32 lanes (overlapping the
//      already-counted ones) and masking the movemask fixes that with no
//      extra correctness risk: the mask removes exactly the lanes outside
//      [lo, hi), and a lane outside that range can only inflate a count.
{
u32 r = hi - aj;
if (r) {
/* [r2j_ TAILAJ] 尾块窗口改取 [aj, aj+32)（原 [hi-32, hi)）⇒ base 就是本环自己的 aj，
   免掉 hi-32 的重新物化与移位掩码计算；掩码 = 32 项表 g_r2j_kmaskT[r]（r ∈ [1,31]）。
   窗口外车道（rank ≥ hi）的载入字节被掩码屏蔽；rank ≤ N 且 N+32 ≤ MF_SPL ⇒ 不越界。
   逐位等价。来源：本账号 1013b #122903（板面定价 -1.36%）。 */
u32 base = aj, lk = 0u;
u32 kmask = g_r2j_kmaskT[r];
/* [ONEACC] tail: same one-accumulator form. */
__m256i ac=_mm256_subs_epi8(LDU256(sbase+base),qsb[0]);
for(int e=1;e<k-1;e++){
__m256i cv=LDU256(sbase+(size_t)e*MF_SPL+base);
ac=_mm256_max_epi8(ac,_mm256_subs_epi8(cv,qsb[e]));
}
u32 decm = ((u32)_mm256_movemask_epi8(ac)) & kmask;
u32 amb = ((u32)_mm256_movemask_epi8(_mm256_cmpeq_epi8(ac, _mm256_setzero_si256()))) & kmask;
aacc += (u32)__builtin_popcount(decm);
while (amb) {
int t = __builtin_ctz(amb); amb &= amb - 1;
if(KK==8 && !c0 && tie_count<TIE_CAP){
 u32 *e=tie_data+(size_t)tie_count++*TIE_EW;
 e[0]=i;e[1]=base+(u32)t;e[2]=d;
 _mm256_storeu_si256((__m256i*)(e+3),qT);
 continue;
}
const u32 *cp = c0 ? (c0 + (size_t)(base + (u32)t - lo) * g_KS) : (s_pm + (size_t)odd[base + (u32)t] * STRIDE);
__m256i pv = LDU256(cp);
u32 okl = (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
if (k > 8) {
__m256i pvh = LDU256((cp + 8));
okl &= (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
}
aacc += okl;
}
}
}
add = aacc;
} else if (c0) {
if (k <= 8) {
const u32 *c = c0;
u32 a0 = 0, a1 = 0, a2 = 0, a3 = 0;
u32 j = lo;
for (; j + 4 <= hi; j += 4) {
__m256i p0 = LDU256((c));
__m256i p1 = LDU256((c + g_KS));
__m256i p2 = LDU256((c + 2 * g_KS));
__m256i p3 = LDU256((c + 3 * g_KS));
a0 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p0), all8);
a1 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p1), all8);
a2 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p2), all8);
a3 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p3), all8);
c += 4 * g_KS;
}
for (; j < hi; j++, c += g_KS) {
__m256i pv = LDU256(c);
a0 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
}
add = a0 + a1 + a2 + a3;
} else {
// ---- k>8 fringe: unroll 4x.  The k<=8 path below has always been unrolled
//      4x; the k>8 path was left as one row per iteration with a single serial
//      `add +=` chain, even though k=9/10 is exactly where the fringe dominates
//      (measured at k=10/n=1.5e5: fringe 332 ms = 57 % of the run vs AND 139 ms
//      = 24 %).  Four independent accumulators + two loads in flight per row.
const u32 *c = c0;
u32 a0 = 0, a1 = 0, a2 = 0, a3 = 0;
u32 j = lo;
for (; j + 4 <= hi; j += 4) {
__m256i p0 = LDU256((c));
__m256i p1 = LDU256((c + g_KS));
__m256i p2 = LDU256((c + 2 * g_KS));
__m256i p3 = LDU256((c + 3 * g_KS));
__m256i h0 = LDU256((c + 8));
__m256i h1 = LDU256((c + g_KS + 8));
__m256i h2 = LDU256((c + 2 * g_KS + 8));
__m256i h3 = LDU256((c + 3 * g_KS + 8));
a0 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p0), all8) & (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, h0), all8);
a1 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p1), all8) & (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, h1), all8);
a2 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p2), all8) & (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, h2), all8);
a3 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, p3), all8) & (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, h3), all8);
c += 4 * g_KS;
}
for (; j < hi; j++, c += g_KS) {
__m256i pv = LDU256(c);
__m256i pvh = LDU256((c + 8));
a0 += (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8) &
(u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
}
add = a0 + a1 + a2 + a3;
}
} else {
// Fallback used when the transposed fringe rows are not allocated.  NOTE: this
// branch used to test only lanes 0..7, so for k>8 the dims 8..k-1 were never
// compared and the fringe silently OVERCOUNTED (found on the board: k=9 at n=3e5
// returned WA while k=10 happened to pass).  MF_NOSDMP=1 is therefore only safe
// with the high-lane test below; keep MF_NOSDMP=0 by default regardless.
for (u32 j = lo; j < hi; j++) {
const u32 *c = s_pm + (size_t)odd[j] * STRIDE;
__m256i pv = _mm256_load_si256((const __m256i *)c);
u32 ok = (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qT, pv), all8);
if (k > 8) {
__m256i pvh = _mm256_load_si256((const __m256i *)(c + 8));
ok &= (u32)_mm256_testc_si256(_mm256_cmpgt_epi32(qTh, pvh), all8);
}
add += ok;
}
}
out[i] += add;
}
};
if constexpr(KK>0) fringe_dim(std::integral_constant<int,0>{});
if constexpr(KK>1) fringe_dim(std::integral_constant<int,1>{});
if constexpr(KK>2) fringe_dim(std::integral_constant<int,2>{});
if constexpr(KK>3) fringe_dim(std::integral_constant<int,3>{});
if constexpr(KK>4) fringe_dim(std::integral_constant<int,4>{});
if constexpr(KK>5) fringe_dim(std::integral_constant<int,5>{});
if constexpr(KK>6) fringe_dim(std::integral_constant<int,6>{});
if constexpr(KK>7) fringe_dim(std::integral_constant<int,7>{});
if constexpr(KK>8) fringe_dim(std::integral_constant<int,8>{});
if constexpr(KK>9) fringe_dim(std::integral_constant<int,9>{});
}
__attribute__((target("avx2")))
__attribute__((aligned(1024))) static void solve_mf(u32 N, const unsigned **x, u32 *out, int k) {
++solve_generation;   // [对手 #121252] 入口自增：avN 初值 0 ⇒ 首次调用必建（勿移到末尾）
// W padded to 8 words: every (dim,bucket) array then starts on a 64-byte
// boundary and every 32-byte vector load inside a stream is line-contained.
const u32 W = ((N + 63) / 64 + 7u) & ~7u;
g_W = W;
#ifdef MF_SKIP_AND
g_skip_and = 1;
#endif
#ifdef MF_FILT
g_use_filt = 1;
#endif
#ifdef MF_NOFILT
g_use_filt = 0;
#endif
#ifdef MF_Q
g_Q = MF_Q; g_B = (u32)((N + (MF_Q) - 1) / (MF_Q));
#endif
#ifdef MF_FILT
g_use_filt = 1;
#endif
#ifdef MF_NOFILT
g_use_filt = 0;
#endif
#ifdef MF_SKIP_FRINGE
g_skip_fringe = 1;
#endif
if (!g_Q) {
// bucket size ~ 0.82*sqrt(n) balances the pool build (proportional to n^2/B) against the
// fringe scan (proportional to n*B);  the pool itself is capped by the memory budget.
u64 r = (u64)N, s2 = 0;
while (s2 * s2 < r) s2++;                       // ceil(sqrt(n)) without libm
// Reserve: the .bss tables, the coarse-filter bitmaps and (for k>=9, where the
// random s_pm fallback is fatal) the transposed fringe rows.
u64 overhead = (u64)(5 * k + 21) * N * 4;
#if !MF_NOFILTB
// STALE-SHAPE RESERVE (the KG_BASE / CT_MAXV class, on the MEMORY BUDGET rather than a
// buffer).  The coarse-filter bitmaps are k*k*(Qc+1)*W*8 bytes, approximated here as
// k*k*4*N.  But `#if MF_NOFILTB` sets g_use_filt = 0 UNCONDITIONALLY further down, and
// MF_FILTAUTO's own auto-enable runs AFTER this block -- so on a row shipping
// MF_NOFILTB 1 the bitmaps are never allocated on ANY path, and this reserves
// k*k*4*N = 324 MB at n=1e6/k=9 for memory that cannot be used.  The cost is not the
// 324 MB: it is that Qmem = budget/perfill then caps q at 1997 instead of the 2440 this
// engine's own coefficient asks for, i.e. a 22 % coarser grid and ~18 % MORE fringe
// candidates on the phase that is 77 % of the row.
overhead += (u64)k * k * 4 * N;
#endif
// The byte-signature planes are k*k*(N+64) bytes and were NOT charged here, so at
// n=1e6/k=10 the pool was sized ~100 MB too large, the allocation failed, and the
// engine returned WRONG ANSWERS on the board (1016-SIG8, sid 98573, WA at 8.929 s)
// instead of erroring -- base8, which has no planes, was AC on the same row.  Charge
// them, and refuse the signature path outright if the planes alone would take too
// large a share of the budget.
overhead += (u64)k * k * ((u64)N + 64);
overhead += (u64)N * ((k <= 8) ? 8u : (u32)k) * 4 * k;
u64 budget = ((u64)MF_BUDGETMB << 20);
if (budget > overhead + ((u64)64 << 20)) budget -= overhead; else budget = (u64)64 << 20;
// MF_PLEN sizes each pooled array to the largest R ever read from it, so the
// pool really used is only ~fill of (k-1)*(Q+1)*W words; charge that instead.
#if MF_PLEN
static const u32 kfill[7] = {72, 67, 62, 58, 54, 50, 46};   // per-array fill, k=4..10
u64 perfill = ((u64)(k - 1) * W * 8) * kfill[k - 4] / 100;
#else
u64 perfill = (u64)(k - 1) * W * 8;
#endif
u64 Qmem = budget / (perfill ? perfill : 1);
if (Qmem < 8) Qmem = 8;
static const u32 kscale[7] = {82, 91, 100, 108, 115, 122, 129};  // sqrt(k/6) in percent, k=4..10
u32 q = (u32)((s2 * MF_QCOEF + 99) / 100);
q = (u32)(((u64)q * kscale[k - 4] + 99) / 100);
if (k == 9 && N >= 900000) q = 1400;
if ((u64)q > Qmem) q = (u32)Qmem;
if (q > N) q = N;
// MAXQ is a SILENT CLIFF: s_maxr is [MAXK][MAXK][MAXQ] and is indexed with `b <= Q`, so
// Q > MAXQ-1 walks into s_pofs/s_plen with correct-looking output.  Free insurance.
if (q > (u32)(MAXQ - 1)) q = (u32)(MAXQ - 1);
g_Q = q;
g_B = (N + q - 1) / q;
}
const u32 Q = g_Q, B = g_B;
const u64 bmag = ((1ull << 40) + B - 1) / B;
#if MF_PLEN
g_plen_on = 1;   // independent of MF_WTRIM: lengths only, stride stays W
#endif
#if MF_WTRIM
u32 Wpool = W;   // MF_WTRIM is KNOWN-BROKEN, see the warning above
#else
const u32 Wpool = W;
#endif
#if MF_FILTAUTO
if (!g_use_filt && B >= 400) g_use_filt = 1;   // coarse grid: the filter beats the scan
#endif
if (g_use_filt && (u64)k * k * (g_Qc + 1) * W * 8 > ((u64)700 << 20)) g_use_filt = 0;
#if !MF_FILTAUTO
g_use_filt = 0;
#endif
{
int lo8[8], hi8[8];
for (int e = 0; e < 8; e++) lo8[e] = (e < k) ? 0 : 0x7FFFFFFF;
for (int e = 0; e < 8; e++) hi8[e] = (8 + e < k) ? 0 : 0x7FFFFFFF;
khi8v = LDU256(lo8);
khi8hv = LDU256(hi8);
}
const int K1 = k - 1;
// ---------- per-dimension orders, ranks, tie info ----------
for (int d = 0; d < k; d++) {
const unsigned *xd = x[d];
u32 *od = s_ord[d];
// [v4_orders] gt[i] = start[x_d[i]] and bd[] from start[x_d[i]+1]: value-indexed
// lookups, so the per-point work is two sequential sweeps of xd[] with no random
// od[]/xd[] traffic (the old loop walked od[] reading xd[od[j]] at random and then
// scattered gt[]/s_bd[] writes once per tie group).
static u8 hist8[MAXN+1] __attribute__((aligned(64)));
for(u32 i=0;i<=N;i++)hist8[i]=0;
for(u32 i=0;i<N;i++){
  if(i+64<N)_mm_prefetch((const char*)(xd+i+64),_MM_HINT_NTA);
  u32 v=xd[i];hist8[v<N?v:0]++;
}
u32 sum=0;
u32 hv=0;
for(;hv+8<=N+1;hv+=8){
 __m256i c=_mm256_cvtepu8_epi32(_mm_loadl_epi64((const __m128i*)(hist8+hv)));
 __m256i p=_mm256_add_epi32(c,_mm256_slli_si256(c,4));
 p=_mm256_add_epi32(p,_mm256_slli_si256(p,8));
 __m128i lo=_mm256_castsi256_si128(p),hi=_mm256_extracti128_si256(p,1);
 hi=_mm_add_epi32(hi,_mm_shuffle_epi32(lo,255));
 __m256i inc=_mm256_inserti128_si256(_mm256_castsi128_si256(lo),hi,1);
 __m256i ex=_mm256_add_epi32(_mm256_sub_epi32(inc,c),_mm256_set1_epi32(sum));
 _mm256_storeu_si256((__m256i*)(s_tmp+hv),ex);
 sum+=(u32)_mm_extract_epi32(hi,3);
}
for(;hv<=N;hv++){u32 c=hist8[hv];s_tmp[hv]=sum;sum+=c;}
bool compact=(sum==N);
if(!compact){
  for(u32 i=0;i<=N;i++)s_cntv[i]=0;
  for(u32 i=0;i<N;i++){u32 v=xd[i];s_cntv[v<N?v:0]++;}
  sum=0;for(u32 i=0;i<=N;i++){u32 c=s_cntv[i];s_cntv[i]=sum;s_tmp[i]=sum;sum+=c;}
}
u32 *gt=s_gt[d],*pp=s_pos[d];
auto fill=[&](auto tag){
 for(u32 i=0;i<N;i++){
  if(i+MF_PFSC2<N){u32 vp=xd[i+MF_PFSC2];if(vp>=N)vp=0;_mm_prefetch((const char*)(s_tmp+vp),_MM_HINT_T0);}
  if(i+64<N)_mm_prefetch((const char*)(xd+i+64),_MM_HINT_NTA);
  u32 v=xd[i];if(v>=N)v=0;
  u32 st,q,num;
  if constexpr(decltype(tag)::value){u32 w=s_tmp[v];st=w&0xfffffu;q=st+(w>>20);s_tmp[v]=w+(1u<<20);num=(s_tmp[v+1]&0xfffffu)-1u;}
  else{st=s_cntv[v];q=s_tmp[v]++;num=s_cntv[v+1]-1u;}
  _mm_stream_si32((int*)(gt+i),(int)st);_mm_stream_si32((int*)(pp+i),(int)q);
#if MF_FB2
  s_posp[(size_t)i*MAXK+d]=q;
#endif
  if(num>N)num=N;
  u32 bb=(u32)(((u64)num*bmag)>>40);
  s_bdT[(size_t)d*E6Z_BDR+i]=bb>Q-1?Q-1:bb;
 }
};
if(compact)fill(std::true_type{});else fill(std::false_type{});
for(u32 i=0;i<N;i++){
#if MF_PFSC2 > 0
 if(i+64<N)_mm_prefetch((const char*)(pp+i+64),_MM_HINT_NTA);
 if(i+MF_PFSC2<N)_mm_prefetch((const char*)(od+pp[i+MF_PFSC2]),_MM_HINT_T0);
#endif
 od[pp[i]]=i;
}
BVAL(d, 0) = 0; FPOS(d, 0) = 0;
for (u32 b = 1; b < Q; b++) {
u32 pos = b * B < N ? b * B : N - 1;
u32 v = xd[od[pos]];
BVAL(d, b) = v;
u32 gs = pos;
while (gs > 0 && xd[od[gs - 1]] == v) gs--;
FPOS(d, b) = gs;
}
BVAL(d, Q) = 0xFFFFFFFFu; FPOS(d, Q) = 0;
}
_mm_sfence();
// [TRANS] one blocked transpose  s_bdT[d][i] -> s_bd[i][d].
// Reads: 9 sequential streams over the dim-major rows.  Writes: 64 points x 10 lanes
// = 1280 contiguous bytes per block, so every 64-B line of s_bd is written WHOLE
// (no read-for-ownership, no partial writeback) -- what the 9 strided passes cost.
for (u32 i0 = 0; i0 < N; i0 += 64) {
u32 i1 = i0 + 64; if (i1 > N) i1 = N;
for (u32 i = i0; i < i1; i++) {
u16 *dst = s_bd + (size_t)i * MAXK;
for (int dd = 0; dd < k; dd++) dst[dd] = s_bdT[(size_t)dd * E6Z_BDR + i];
}
}
{ u32 nn_ = N - 1; u32 sh_ = 0; while (nn_ > 255u) { nn_ >>= 1; sh_++; } g_sigsh = sh_; }
const int b14af_fill_sig = (k == STRIDE) && ((size_t)k * k * MF_SPL <= ((size_t)220 << 20)) && ((size_t)N + 32 <= MF_SPL);
if (k == STRIDE) {
for (u32 i = 0; i < N; i++) {
u32 *p = s_pm + (size_t)i * STRIDE;
__m256i pv = _mm256_setr_epi32((int)x[0][i],(int)x[1][i],(int)x[2][i],(int)x[3][i],
(int)x[4][i],(int)x[5][i],(int)x[6][i],(int)x[7][i]);
_mm256_stream_si256((__m256i *)p, pv);
if (b14af_fill_sig) { u8 *cs_ = b14af_csig + (size_t)i * 8;
for (int e = 0; e < 8; e++) cs_[e] = (u8)((x[e][i] >> g_sigsh) ^ 0x80u); }
}
_mm_sfence();
} else {
for (u32 i = 0; i < N; i++) {
u32 *p = s_pm + (size_t)i * STRIDE;
for (int d = 0; d < k; d++) p[d] = x[d][i];
for (int d = k; d < STRIDE; d++) p[d] = 0;
}
}
// ---------- byte-signature planes for the fringe coarse filter ----------
// s_sig[d][e*N + j] = (u8)(x[e][ od_d[j] ] >> SIGSH)   (od_d = dim-d order)
{
u32 nn = N - 1; u32 sh = 0;
while (nn > 255u) { nn >>= 1; sh++; }        // smallest sh with (N-1)>>sh <= 255
g_sigsh = sh;
size_t need = (size_t)k * (k-1) * MF_SPL;
static u8 *sigpool = 0; static size_t sigcap = 0;
// Refuse the signature path when the planes alone exceed 220 MB: past that the pool
// it competes with is squeezed and the row is better served by the exact path.
if (need > ((size_t)220 << 20) || (size_t)N + 32 > MF_SPL) { sigpool = 0; sigcap = 0; }
else if (need > sigcap) { sigpool = (u8 *)pool_alloc(need); sigcap = sigpool ? need : 0; }
if (sigpool) {
s_sig = sigpool;
// The plane build is DEFERRED and merged into the s_sdmp build pass below when
// s_sdmp is allocated: both walk the same `s_pm + od[j]*STRIDE` rows, so two
// passes pay two rounds of random 64-byte reads for one round of useful work.
// ⚠ SAFETY GATE.  The signature path is VERIFIED CORRECT (brute force k=4..10 x 4
// distributions, plus checksum-identical to the base engine) for SMALL bucket sizes,
// but at large B it returns WRONG ANSWERS -- reproduced twice each way, real judge:
//   k=6 n=261000  B=326 (QCOEF 800)  checksum 1070791438 = base engine  CORRECT
//   k=6 n=261000  B=428 (QCOEF 610)  checksum 1064212397 != base          WRONG
//   k=6 n=261000  B=652 (QCOEF 400)  checksum 1060739010 != base          WRONG
// and the board agrees: 1012b/1013b/1014b (n=1e5, B=259) AC; 1013a (n=3e5, B=416)
// and 1016 (n=1e6, B=842) both WA.  The threshold sits between B=326 and B=428.
// Root cause NOT yet localised (lane-d neutralisation, the masked tail, and the
// exact-tail variant were each ruled out by checksum -- all give the same wrong
// value, so the defect is in the coarse scan or the planes themselves).
// UNTIL IT IS FOUND, REFUSE THE SIGNATURE PATH ABOVE A MEASURED-SAFE B.
g_sig_on = 1;
g_sig_merged = 0;
}
}
// ---------- coarse superset bitmaps over every dim order (fringe filter) ----------
#if MF_NOFILTB
g_use_filt = 0;
#endif
#if MF_FB3
if (g_use_filt) {
u32 Qc = g_Qc;
u64 wc = (u64)N / Qc + 1;
for (int e = 0; e < k; e++) {
const unsigned *xe = x[e];
const u32 *pe = s_ord[e];
u32 ptr = 0;
u32 *bn = s_bnde + (size_t)e * (MAXQ + 2);
for (u32 c = 0; c <= Qc; c++) {
u64 v = (u64)c * wc;
if (v > 0xFFFFFFFFull) v = 0xFFFFFFFFull;
while (v && ptr < N && (u64)xe[pe[ptr]] < v) ptr++;
bn[c] = ptr;
}
}
}
#endif
if (g_use_filt) {
u32 Qc = g_Qc;
u64 wc = (u64)N / Qc + 1;
size_t need2 = (size_t)k * k * (Qc + 1) * W * sizeof(u64);
static u64 *fpool = 0; static size_t fcap = 0;
if (need2 > fcap) { fpool = (u64 *)pool_alloc(need2 + 4096); fcap = need2; }
if (fpool) {
s_fbm = fpool; g_fbm_W = W;
#if MF_FB2
// Sweep the e-order ONCE per dim e, keeping (k-1) running bitsets (one per
// target dim d), and copy them out at each coarse-cell boundary.  The old
// shape re-walked all N points for every one of the k(k-1) (d,e) pairs with
// two L3-random reads each (xe[pe[ptr]] and s_pos[d][pe[ptr]]); here the
// rank record is point-major so one 64B line yields every dim's rank, which
// cuts the random reads by ~k and the whole build by ~2-2.5x at k>=8.
for (int d = 0; d < k; d++)
for (int e = 0; e < k; e++) {
if (e == d) continue;
u64 *b0 = s_fbm + (size_t)(d * k + e) * (Qc + 1) * W;
for (u32 w = 0; w < W; w++) b0[w] = 0;
}
for (int e = 0; e < k; e++) {
const unsigned *xe = x[e];
const u32 *pe = s_ord[e];
int dmap[MAXK], km1 = 0;
for (int d = 0; d < k; d++) if (d != e) dmap[km1++] = d;
for (int i = 0; i < km1; i++) { u64 *a = s_facc + (size_t)i * W; for (u32 w = 0; w < W; w++) a[w] = 0; }
u32 ptr = 0;
for (u32 c = 1; c <= Qc; c++) {
u64 v = (u64)c * wc;
if (v > 0xFFFFFFFFull) v = 0xFFFFFFFFull;
while (ptr < N && (u64)xe[pe[ptr]] < v) {
const u32 *pv = s_posp + (size_t)pe[ptr] * MAXK;
for (int i = 0; i < km1; i++) {
u32 q = pv[dmap[i]];
s_facc[(size_t)i * W + (q >> 6)] |= 1ull << (q & 63);
}
ptr++;
}
for (int i = 0; i < km1; i++) {
int d = dmap[i];
u64 *dst = s_fbm + ((size_t)(d * k + e) * (Qc + 1) + c) * W;
const u64 *src = s_facc + (size_t)i * W;
#if MF_NT
{ u32 w = 0;
for (; w + 4 <= W; w += 4) _mm256_stream_si256((__m256i *)(dst + w), _mm256_load_si256((const __m256i *)(src + w)));
for (; w < W; w++) dst[w] = src[w]; }
#else
for (u32 w = 0; w < W; w++) dst[w] = src[w];
#endif
}
}
}
#else
for (int d = 0; d < k; d++) {
for (int e = 0; e < k; e++) {
if (e == d) continue;
const unsigned *xe = x[e];
const u32 *pe = s_ord[e];
const u32 *pd = s_pos[d];
u64 *base = s_fbm + (size_t)(d * k + e) * (Qc + 1) * W;
for (u32 w = 0; w < W; w++) base[w] = 0;
u64 *cur = s_acc;
for (u32 w = 0; w < W; w++) cur[w] = 0;
u32 ptr = 0;
for (u32 c = 1; c <= Qc; c++) {
#if MF_FB3
u32 pend = s_bnde[(size_t)e * (MAXQ + 2) + c];
#else
u64 v = (u64)c * wc;
if (v > 0xFFFFFFFFull) v = 0xFFFFFFFFull;
u32 pend = ptr;
{ const unsigned *xe_ = xe; const u32 *pe_ = pe;
while (pend < N && (u64)xe_[pe_[pend]] < v) pend++; }
#endif
while (ptr < pend) { u32 q = pd[pe[ptr]]; cur[q >> 6] |= 1ull << (q & 63); ptr++; }
u64 *dst = base + (size_t)c * W;
#if MF_NT
{ u32 w = 0;
for (; w + 4 <= W; w += 4) _mm256_stream_si256((__m256i *)(dst + w), _mm256_load_si256((const __m256i *)(cur + w)));
for (; w < W; w++) dst[w] = cur[w]; }
#else
for (u32 w = 0; w < W; w++) dst[w] = cur[w];
#endif
}
}
}
#endif
} else { g_use_filt = 0; }
_mm_sfence();
}
// ---------- fold dimension + bucket vector per query ----------
{
size_t mrn = (size_t)MAXK * MAXK * MAXQ;
for (size_t z = 0; z < mrn; z++) s_maxr[z] = 0;
}
// [FOLDBLK] the argmin pass and the maxr-reduction pass are split and blocked: the
// reduction's table is only 252 KB (7 ds x 7 d x 1512 b) but the argmin streams 28 MB of
// s_gt through L2 right next to it, so every one of the 6e6 conditional stores paid an
// L3/DRAM round trip.  Blocking keeps the landing zone resident; the fold result is
// bit-identical (same max reduction, same values, same s_g).
{
static u32 y2_dsb[1024], y2_bst[1024];
for (u32 i0 = 0; i0 < N; i0 += 1024) {
u32 i1 = i0 + 1024; if (i1 > N) i1 = N;
for (u32 i = i0; i < i1; i++) {
u32 best = 0xFFFFFFFFu; int ds = 0;
for (int d = 0; d < k; d++) {
u32 g = s_gt[d][i];
if (g < best) { best = g; ds = d; }
}
y2_dsb[i - i0] = (u32)ds;
y2_bst[i - i0] = best;
s_g[i] = (u8)ds;
}
if (g_plen_on) {
// Largest R that will ever be read out of each (fold dim, dim, bucket) array:
// the pooled array (ds,d,b) is only ever scanned over [0, maxR] words, so the
// per-array length can be far below the group maximum (0.38n vs 0.93n at k=10).
for (u32 i = i0; i < i1; i++) {
u32 ds = y2_dsb[i - i0];
u32 bb = y2_bst[i - i0];
const u16 *bdv = s_bd + (size_t)i * MAXK;
u32 *mr = s_maxr + ((size_t)ds * MAXK) * MAXQ;
for (int d = 0; d < k; d++) {
if (d == ds) continue;
u32 b = bdv[d];
/* [b14af_ MAXRBL] 原式是「条件写」，7 条/点的数据相关分支（`bb` 是随机秩、
   单元格最大值在块内递增）⇒ 预测器吃不消。改成**无条件取最大后写回**（同值、逐位同），
   让 gcc 发 cmov ⇒ 无分支。 */
{ u32 *mp = &mr[(size_t)d * MAXQ + b]; u32 v = *mp; *mp = (bb > v) ? bb : v; }
}
}
}
}
}
// group by ds
{
for (u32 v = 0; v <= (u32)k; v++) s_cntv[v] = 0;
for (u32 i = 0; i < N; i++) s_cntv[s_g[i] + 1]++;
{ u32 ac = 0; for (u32 v = 0; v <= (u32)k; v++) { u32 c = s_cntv[v]; s_cntv[v] = ac; ac += c; } }
for (u32 i = 0; i < N; i++) { s_tmp[s_cntv[s_g[i] + 1]++] = i; }
for (u32 i = 0; i < N; i++) s_qord[i] = s_tmp[i];
}
u32 gstart[MAXK + 2];
{ u32 c = 0; for (int d = 0; d <= k; d++) { gstart[d] = c; while (c < N && (int)s_g[s_qord[c]] == d) c++; } gstart[k + 1] = c; }
// [FRORD] Reuse each global coordinate order, partitioning once by fold group.
// The preexisting s_bdT holds each point's exact bucket in a compact column.
for (int d = 0; d < k; d++) {
u32 pos[MAXK];
for (int ds = 0; ds < k; ds++) pos[ds] = gstart[ds];
u32 bcap_ = (N - 1) / B; if (bcap_ > Q - 1) bcap_ = Q - 1;
u32 bb_ = 0;
u32 rnext = (bcap_ >= 1u) ? FPOS(d, 1u) : 0xFFFFFFFFu;
for (u32 r = 0; r < N; r++) {
if (r >= rnext) { do { bb_++; rnext = (bb_ < bcap_) ? FPOS(d, bb_ + 1u) : 0xFFFFFFFFu; } while (r >= rnext); }
#if MF_FRPF > 0
/* [b14ae_ FRPF / N1 律] 本环 = 顺序 `s_ord[d][r]` -> **随机 gather `s_g[id]`**（4 MB 表）
   -> 8 游标散布写 `s_frord[d]`。本环实测 ~33 拍/迭代（本席相位仪）⇒ 远 ≥ 8 拍 ⇒ 按 N1
   定标律正是该加**链式目标预取**的地方（与 #123047/#123142 在池 OR 环兑现的同一手法）：
   提前 MF_FRPF 个 r 把第二级地址 `s_g[s_ord[d][r+K]]` 算出来并发出 T0 预取。纯预取，逐位等价。 */
if (r + MF_FRPF < N) _mm_prefetch((const char *)(s_g + s_ord[d][r + MF_FRPF]), _MM_HINT_T0);
#endif
u32 id = s_ord[d][r];
u32 dst = pos[s_g[id]]++;
s_frord[d][dst] = id | (bb_ << 20);
}
}
// ---------- transposed coordinate rows (sequential fringe) ----------
#if MF_NOSDMP
g_nosdmp = 1;
#endif
if (!g_nosdmp) {
u32 KS = k;
#if MF_KSPAD
if (k <= 8) KS = 8;          // 32-byte aligned candidate rows for k<=8
#endif
size_t per = (size_t)N * KS * 4;
size_t lim = (size_t)150 << 20;
if (k >= 9) lim = (size_t)470 << 20;   // k=9/10 need the rows: the random s_pm fallback is fatal
if (per * k <= lim) {
static u32 *pool = 0; static size_t cap = 0;
if (per * k > cap) { pool = (u32 *)pool_alloc(per * k + 256); cap = per * k; }
if (pool) {
g_KS = KS;
for (int d = 0; d < k; d++) {
s_sdmp[d] = pool + (size_t)d * N * g_KS;
const u32 *od = s_ord[d];
u32 *dst = s_sdmp[d];
const u32 KSx = g_KS;
for (u32 j = 0; j < N; j++) {
#if MF_PFB > 0
if (j + MF_PFB < N) _mm_prefetch((const char *)(s_pm + (size_t)od[j + MF_PFB] * STRIDE), _MM_HINT_T0);
#endif
const u32 *row = s_pm + (size_t)od[j] * STRIDE;
u32 *dp = dst + (size_t)j * KSx;
for (u32 e = 0; e < KSx; e++) dp[e] = (e < (u32)k) ? row[e] : 0u;
if (s_sig) {
u8 *sb = s_sig + (size_t)d * (k-1) * MF_SPL;
for(int e=0;e<k;e++)if(e!=d){u32 z_=row[e]>>g_sigsh;sb[(size_t)(e-(e>d))*MF_SPL+j]=(u8)(z_^0x80u);}
}
}
if (s_sig) g_sig_merged = 1;
}
}
}
}
// ---------- signature planes, standalone only if the s_sdmp pass did not run ----
if (s_sig && !g_sig_merged && k==8) {
const u8 *const cached_signature = b14af_csig;
for(int d=0;d<k;d++) {
const u32 *od=s_ord[d];
u8 *base=s_sig+(size_t)d*(k-1)*MF_SPL;
u32 j=0;
auto emit=[&](auto tag){constexpr int OMIT=decltype(tag)::value;
alignas(64) u8 pbuf[8][64]; u32 jc = 0, jg_ = 0;
for(;j+8<=N;j+=8) {
if (jc == 0) jg_ = j;
/* [b14ae_ SIGPF] 本环每次迭代要 **8 条随机 8 B 行 gather**（cached_signature[od[j+e]*8]，
   e=0..7），而基线只给 e=0 那条发了 T0 预取 ⇒ 其余 7 条全裸奔。按 N1 律把**八条全发**：
   一次把下一轮（距离 32 = 4 轮）的 8 个目标行都点上。纯预取，逐位等价。 */
if(j+128<N)for(int e=0;e<8;e++)_mm_prefetch((const char *)(cached_signature+(size_t)od[j+128+e]*8),_MM_HINT_T0);
__m128i r0=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+0]*8));
__m128i r1=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+1]*8));
__m128i r2=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+2]*8));
__m128i r3=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+3]*8));
__m128i r4=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+4]*8));
__m128i r5=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+5]*8));
__m128i r6=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+6]*8));
__m128i r7=_mm_loadl_epi64((const __m128i *)(cached_signature+(size_t)od[j+7]*8));
__m128i t0=_mm_unpacklo_epi8(r0,r1);
__m128i t1=_mm_unpacklo_epi8(r2,r3);
__m128i t2=_mm_unpacklo_epi8(r4,r5);
__m128i t3=_mm_unpacklo_epi8(r6,r7);
__m128i u0=_mm_unpacklo_epi16(t0,t1);
__m128i u1=_mm_unpackhi_epi16(t0,t1);
__m128i u2=_mm_unpacklo_epi16(t2,t3);
__m128i u3=_mm_unpackhi_epi16(t2,t3);
__m128i v0=_mm_unpacklo_epi32(u0,u2);
__m128i v1=_mm_unpackhi_epi32(u0,u2);
__m128i v2=_mm_unpacklo_epi32(u1,u3);
__m128i v3=_mm_unpackhi_epi32(u1,u3);
__m128i h0=_mm_unpackhi_epi8(r0,r1);
__m128i h1=_mm_unpackhi_epi8(r2,r3);
__m128i h2=_mm_unpackhi_epi8(r4,r5);
__m128i h3=_mm_unpackhi_epi8(r6,r7);
__m128i hu0=_mm_unpacklo_epi16(h0,h1),hu1=_mm_unpacklo_epi16(h2,h3);
__m128i v4=_mm_unpacklo_epi32(hu0,hu1);
if constexpr(OMIT!=0)_mm_storel_epi64((__m128i*)(pbuf[(0-(0>OMIT))]+jc*8),v0);
if constexpr(OMIT!=1)_mm_storel_epi64((__m128i*)(pbuf[(1-(1>OMIT))]+jc*8),_mm_srli_si128(v0,8));
if constexpr(OMIT!=2)_mm_storel_epi64((__m128i*)(pbuf[(2-(2>OMIT))]+jc*8),v1);
if constexpr(OMIT!=3)_mm_storel_epi64((__m128i*)(pbuf[(3-(3>OMIT))]+jc*8),_mm_srli_si128(v1,8));
if constexpr(OMIT!=4)_mm_storel_epi64((__m128i*)(pbuf[(4-(4>OMIT))]+jc*8),v2);
if constexpr(OMIT!=5)_mm_storel_epi64((__m128i*)(pbuf[(5-(5>OMIT))]+jc*8),_mm_srli_si128(v2,8));
if constexpr(OMIT!=6)_mm_storel_epi64((__m128i*)(pbuf[(6-(6>OMIT))]+jc*8),v3);
if constexpr(OMIT!=7)_mm_storel_epi64((__m128i*)(pbuf[(7-(7>OMIT))]+jc*8),_mm_srli_si128(v3,8));
/* [r2j_ PLANE-NT] 每 8 轮（64 点）把 7 张平面各 64 B 用 2 条 32 B 流存（NT）整行写出：
   写一次、晚读大量（fringe 分类器）⇒ 免掉部分行写的 RFO 读。目标地址 32 B 对齐
   （s_sig 2 MB 对齐、MF_SPL=245*4096、jg_ 为 64 的倍数）。逐位等价。 */
if (++jc == 8) {
jc = 0;
#define R2J_FLUSH_PLANE(E, BUFIDX) if constexpr(OMIT!=E) { __m256i _a=LDU256(pbuf[(BUFIDX)]), _b=LDU256(pbuf[(BUFIDX)]+32); \
  _mm256_stream_si256((__m256i*)(base+(size_t)(E-(E>OMIT))*MF_SPL+jg_),_a); \
  _mm256_stream_si256((__m256i*)(base+(size_t)(E-(E>OMIT))*MF_SPL+jg_+32),_b); }
R2J_FLUSH_PLANE(0,0-(0>OMIT)) R2J_FLUSH_PLANE(1,1-(1>OMIT)) R2J_FLUSH_PLANE(2,2-(2>OMIT)) R2J_FLUSH_PLANE(3,3-(3>OMIT))
R2J_FLUSH_PLANE(4,4-(4>OMIT)) R2J_FLUSH_PLANE(5,5-(5>OMIT)) R2J_FLUSH_PLANE(6,6-(6>OMIT)) R2J_FLUSH_PLANE(7,7-(7>OMIT))
#undef R2J_FLUSH_PLANE
}
}
/* 残余（jc 个 8 点块）用普通存补写；随后 sfence 保证 NT 数据对本核后续读取可见。 */
for (u32 t2 = 0; t2 < jc; t2++) {
#define R2J_RES_PLANE(E, BUFIDX) if constexpr(OMIT!=E) _mm_storel_epi64((__m128i*)(base+(size_t)(E-(E>OMIT))*MF_SPL+(j-8*(jc-t2))+0),_mm_loadl_epi64((const __m128i*)(pbuf[(BUFIDX)]+t2*8)));
R2J_RES_PLANE(0,0-(0>OMIT)) R2J_RES_PLANE(1,1-(1>OMIT)) R2J_RES_PLANE(2,2-(2>OMIT)) R2J_RES_PLANE(3,3-(3>OMIT))
R2J_RES_PLANE(4,4-(4>OMIT)) R2J_RES_PLANE(5,5-(5>OMIT)) R2J_RES_PLANE(6,6-(6>OMIT)) R2J_RES_PLANE(7,7-(7>OMIT))
#undef R2J_RES_PLANE
}
};
switch(d){case 0:emit(std::integral_constant<int,0>{});break;case 1:emit(std::integral_constant<int,1>{});break;case 2:emit(std::integral_constant<int,2>{});break;case 3:emit(std::integral_constant<int,3>{});break;case 4:emit(std::integral_constant<int,4>{});break;case 5:emit(std::integral_constant<int,5>{});break;case 6:emit(std::integral_constant<int,6>{});break;case 7:emit(std::integral_constant<int,7>{});break;}
for(;j<N;j++) {
const u8 *row=cached_signature+(size_t)od[j]*8;
for(int e=0;e<k;e++)if(e!=d)base[(size_t)(e-(e>d))*MF_SPL+j]=row[e];
}
}
_mm_sfence();
g_sig_merged=1;
}
if (s_sig && !g_sig_merged) {
for (int d = 0; d < k; d++) {
const u32 *od = s_ord[d];
u8 *base = s_sig + (size_t)d * (k-1) * MF_SPL;
for (u32 j = 0; j < N; j++) {
#if MF_PFB > 0
if (j + MF_PFB < N) _mm_prefetch((const char *)(s_pm + (size_t)od[j + MF_PFB] * STRIDE), _MM_HINT_T0);
#endif
const u32 *row = s_pm + (size_t)od[j] * STRIDE;
for(int e=0;e<k;e++)if(e!=d){u32 z_=row[e]>>g_sigsh;base[(size_t)(e-(e>d))*MF_SPL+j]=(u8)(z_^0x80u);}
}
}
}
// ---------- pool ----------
u64 poolwords = (u64)K1 * (Q + 1) * W;
#if MF_PLEN
if (g_plen_on) {
u64 best = 0;
for (int dsx = 0; dsx < k; dsx++) {
u64 tot = 0;
for (int e = 0; e < K1; e++) {
int d = (e < dsx) ? e : e + 1;
const u32 *mr = s_maxr + ((size_t)dsx * MAXK + d) * MAXQ;
for (u32 b = 0; b <= Q; b++) {
u32 RR = mr[b];
u32 len = ((RR >> 6) + 8u) & ~7u;
if (len > W) len = W;
tot += len;
}
}
if (tot > best) best = tot;
}
if (best && best < poolwords) poolwords = best;
}
#endif
size_t need = (size_t)poolwords * sizeof(u64);
static u64 *pool = 0; static size_t cap2 = 0;
if (need > cap2) { pool = (u64 *)pool_alloc(need + 4096); cap2 = need; }
s_bs = pool;
#if MF_INCA
/* [INCA-S] 定扎车道顺序流：每群一条、共 k 条（k-1 条共享 d1=0 的走查序）。
★ 在本 block **之外**先算好 budget/Q（见上），这里只分配，不参与任何 Qmem/B 计算。 */
{
size_t needho = (size_t)N * sizeof(u32) + 4096;
if (needho > s_ho_cap) { s_ho = (u32 *)pool_alloc(needho); s_ho_cap = s_ho ? needho : 0; }
g_inca = (s_ho && (u64)N <= (u64)AS_WORDS * 64ull) ? 1 : 0;
}
#endif
// ---------- groups ----------
for (int ds = 0; ds < k; ds++) {
u32 q0 = gstart[ds], q1 = gstart[ds + 1];
if (q0 >= q1) continue;
#if MF_WTRIM
{
u32 mx = 1;
for (u32 qi = q0; qi < q1; qi++) { u32 rr = s_gt[ds][s_qord[qi]]; if (rr > mx) mx = rr; }
u32 wp = ((mx >> 6) + 8u) & ~7u;
Wpool = (wp < W) ? wp : W;
}
#endif
#if MF_PLEN
// Per-(dim,bucket) array lengths inside the group.
if (g_plen_on) {
u64 tot = 0;
for (int e = 0; e < K1; e++) {
int d = (e < ds) ? e : e + 1;
const u32 *mr = s_maxr + ((size_t)ds * MAXK + d) * MAXQ;
u32 *po = s_pofs + (size_t)e * MAXQ;
u32 *pl = s_plen + (size_t)e * MAXQ;
for (u32 b = 0; b <= Q; b++) {
// NOTE: bucket 0 is the EMPTY prefix set (all-zero array), but a query whose
// bd[d]==0 still scans [0,R) words of it, so the array must be long enough --
// sizing it to the fixed 8-word minimum reads past the zeroed region into the
// next array.  Use the real per-bucket max R here, not 0.
u32 RR = mr[b];
u32 len = ((RR >> 6) + 8u) & ~7u;
if (len > Wpool) len = Wpool;
po[b] = (u32)tot; pl[b] = len; tot += len;
}
}
if (tot > (u64)K1 * (Q + 1) * W) { g_plen_on = 0; }   // never; safety
}
#endif
// [y4w NIDFUSE] bit numbering = rank in dim ds.  s_pos[ds] is exactly that inverse
// permutation (filled once per dim in the order phase); the per-group N-random-store
// rebuild that used to happen here is redundant (judge price of the same knife on the
// 1014a/1015a siblings: 22.34 Mc ~= 5.3 ms there; here k*N = 9e6 stores).
const u32 *const spp = s_pos[ds];
for (int e = 0; e < K1; e++) {
int d = (e < ds) ? e : e + 1;
const unsigned *xd = x[d];
const u32 *od = s_ord[d];
#if MF_INCA
/* [INCA-S] 定扎车道（d == d1）：该 fine 平面的**唯一**消费者是 pass A 的 ASCR 整面拷贝
（pass B 只读 s_fbm/s_sdmp/s_pm/s_gt/s_bd/avpool；`s_bs` 的读点全在 query_group 的
pass A 与 PF_DIST 预取里），故这里改为：**不落盘**，把这条走查的取数（spp[od[j]]，
本来就要做）按同一顺序写进顺序流。第 j 个值 ≡ 原走查第 j 次 OR 进 cur 的那个 q。 */
if (g_inca && d == ((ds == 0) ? 1 : 0)) {
u32 *dstho = s_ho;
const u32 *hp = spp;
for (u32 j = 0; j < N; j++) {
if (j + 64 < N) _mm_prefetch((const char *)(od + j + 64), _MM_HINT_NTA);
_mm_stream_si32((int *)&dstho[j], (int)hp[od[j]]); }
_mm_sfence();
continue;
}
#endif
u64 *base = g_plen_on ? s_bs : (s_bs + (size_t)e * (Q + 1) * Wpool);
u64 *cur = s_acc;
for (u32 w = 0; w < Wpool; w++) cur[w] = 0;
if (g_plen_on) {
// bucket 0 is the empty prefix set; bd[] can be 0 so it must be materialised as zeros
u64 *d0 = s_bs + s_pofs[(size_t)e * MAXQ];
u32 l0 = s_plen[(size_t)e * MAXQ];
for (u32 w = 0; w < l0; w++) d0[w] = 0;
}
u32 ptr = 0;
#if MF_BUILD_NOPOOL
while (ptr < N) ptr++;
#else
for (u32 b = 1; b <= Q; b++) {
// {q : x_d[q] < BVAL(d,b)} is EXACTLY the prefix [0,FPOS(d,b)) of the dim-d order
// (FPOS is the tie-group start containing rank b*B), so the per-point value compare
// -- one random xd read per point per (ds,d) pair, k(k-1)N of them -- is redundant.
// b==Q means "every point" and FPOS(d,Q) is the 0 sentinel, hence N.
// Credited to agent a7f799ceb7e63b0cd (worth 4-5% at k=4/5 on the board).
u32 pend = (b < Q) ? FPOS(d, b) : (u32)N;
while (ptr < pend) {
/* [b14ac_ ORPF] the loop chain is ptr -> od[ptr] -> spp[od[ptr]] (a random
   4 MB read) -> RMW into cur.  The index stream od[] is sequential, so the
   value load can be issued K iterations early; prefetch only, bit-identical. */
u32 nx = ptr + 32u; if (nx >= N) nx = ptr;
_mm_prefetch((const char *)&spp[od[nx]], _MM_HINT_T0);
_mm_prefetch((const char *)(od + ptr + 64), _MM_HINT_NTA);
u32 q = spp[od[ptr]]; cur[q >> 6] |= 1ull << (q & 63); ptr++; }
u64 *dst = g_plen_on ? (base + s_pofs[(size_t)e * MAXQ + b])
: (base + (size_t)b * Wpool);
u32 wlen = g_plen_on ? s_plen[(size_t)e * MAXQ + b] : Wpool;
#if MF_NT
// The pool is written once and read 89 GB later, so a plain copy pays a
// read-for-ownership DRAM read per line for nothing.  NT stores remove it
// (measured: this loop was 4.58 GB / ~1.7 s, i.e. 2.7 GB/s, vs 12.5 GB/s
// for a plain memset).
{
u32 w = 0;
for (; w + 4 <= wlen; w += 4)
_mm256_stream_si256((__m256i *)(dst + w), _mm256_load_si256((const __m256i *)(cur + w)));
for (; w < wlen; w++) dst[w] = cur[w];
}
#else
for (u32 w = 0; w < wlen; w++) dst[w] = cur[w];
#endif
}
_mm_sfence();
#endif
}
tie_count=0;
query_group<8>(N, out, ds, q0, q1, Q, Wpool);
flush_ties(out);
}
}
void count_8d(int n, const unsigned *x[8], unsigned *out) { solve_mf((u32)n, (const unsigned **)x, out, 8); }
#ifdef LOCAL_TEST
static void brute_nd(int n, const u32 **x, u32 *out, int k) {
for (int i = 0; i < n; i++) {
u32 c = 0;
for (int j = 0; j < n; j++) {
int ok = 1;
for (int d = 0; d < k; d++) if (!(x[d][j] < x[d][i])) { ok = 0; break; }
c += ok;
}
out[i] = c;
}
}
int main(int argc, char **argv) {
int k = (argc > 1) ? atoi(argv[1]) : 6;
int n = (argc > 2) ? atoi(argv[2]) : 200;
int seed = (argc > 3) ? atoi(argv[3]) : 1;
int mode = (argc > 4) ? atoi(argv[4]) : 0;
int qq = (argc > 5 && atoi(argv[5]) > 0) ? atoi(argv[5]) : 0;
g_Q = qq ? (u32)qq : (u32)(n / 250 + 1);
g_B = (u32)((n + g_Q - 1) / g_Q);
if (argc > 6) g_nosort = atoi(argv[6]);
if (argc > 7) g_nosdmp = atoi(argv[7]);
if (argc > 8) g_skip_and = atoi(argv[8]);
if (argc > 9) g_skip_fringe = atoi(argv[9]);
if (argc > 10) g_use_filt = atoi(argv[10]);
if (argc > 11) g_lexsort = atoi(argv[11]);
const int dm = (mode >= 10) ? mode - 10 : mode;
srand(seed);
static u32 *xs[10];
u32 *out = (u32 *)malloc(sizeof(u32) * n);
u32 *ref = (u32 *)malloc(sizeof(u32) * n);
for (int d = 0; d < k; d++) {
xs[d] = (u32 *)malloc(sizeof(u32) * n);
for (int i = 0; i < n; i++) {
if (dm == 0) xs[d][i] = (u32)(rand() % n);
else if (dm == 1) xs[d][i] = (u32)(rand() % 3);
else if (dm == 2) xs[d][i] = 0;
else if (dm == 3) xs[d][i] = (u32)(n - 1 - (rand() % n));
else if (dm == 4) xs[d][i] = (u32)(i);
else xs[d][i] = (u32)(rand() % n);
}
}
if (mode >= 10) {
struct timespec t0, t1;
clock_gettime(CLOCK_MONOTONIC, &t0);
switch (k) {
case 4: count_4d(n, (const unsigned **)xs, out); break;
case 5: count_5d(n, (const unsigned **)xs, out); break;
case 6: count_6d(n, (const unsigned **)xs, out); break;
case 7: count_7d(n, (const unsigned **)xs, out); break;
case 8: count_8d(n, (const unsigned **)xs, out); break;
case 9: count_9d(n, (const unsigned **)xs, out); break;
default: count_10d(n, (const unsigned **)xs, out); break;
}
clock_gettime(CLOCK_MONOTONIC, &t1);
double ms = (t1.tv_sec - t0.tv_sec) * 1e3 + (t1.tv_nsec - t0.tv_nsec) / 1e6;
u64 s = 0; for (int i = 0; i < n; i++) s += out[i];
printf("k=%d n=%d Q=%u TIME %.1f ms  checksum %llu\n", k, n, g_Q, ms, s);
return 0;
}
switch (k) {
case 4: count_4d(n, (const unsigned **)xs, out); break;
case 5: count_5d(n, (const unsigned **)xs, out); break;
case 6: count_6d(n, (const unsigned **)xs, out); break;
case 7: count_7d(n, (const unsigned **)xs, out); break;
case 8: count_8d(n, (const unsigned **)xs, out); break;
case 9: count_9d(n, (const unsigned **)xs, out); break;
default: count_10d(n, (const unsigned **)xs, out); break;
}
brute_nd(n, (const u32 **)xs, ref, k);
int bad = 0;
for (int i = 0; i < n; i++) if (out[i] != ref[i]) { if (bad < 5) printf("MISMATCH i=%d got=%u ref=%u\n", i, out[i], ref[i]); bad++; }
printf("k=%d n=%d mode=%d Q=%u %s (%d mismatches)\n", k, n, mode, g_Q, bad ? "FAIL" : "OK", bad);
return bad != 0;
}
#endif

评测结果