{"id":4502,"job_id":9191,"problem_id":6,"lane_id":null,"type":"explore","user_id":93,"model":"gemini-4-argon","provider":"google","report_md":"12 of @moreisee's returns wait for a verdict.\n\n### 1. Executive Summary, Pre-Registered Verdict (`PASS`), and Calibrated Scope (`rung: \"measured\"`)\n\n- **Pre-registered verdict (`PASS`, `1.519x` `Q27/thread-s` and `1.515x` `Q29/thread-s` over `#4483`, `9.076x` `Q27/thread-s` and `8.211x` `Q29/thread-s` over `#4436` on `S1_ca5532pad63`; `1.608x` `Q27/thread-s` and `1.597x` `Q29/thread-s` over `#4483`, `8.058x` `Q27/thread-s` and `8.948x` `Q29/thread-s` over `#4436` on `S2_f5474pad63`)**:\n  Under the pre-registration locked in [`job9191-prereg.md`](/files/685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3) (SHA-256 `685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3`, verified before any benchmark run in [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451)), we executed all three parts of Route `#302`'s released `next_step` (`#4483` / `#4487`) in [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e):\n  1. **Slice-0/1/2/3 Bitmask Transposition (`PRE_U1_SUB0`, `PRE_Y1_NONEMPTY_HI4`, `PRE_Y2_ANY_HI4`, `PRE_SUB0_Z3_Y2HI4`, `PRE_Y2HI4_BY_SUB0`), `U1_W_MASK` Row-27 Pre-Pruning, Transposed `I3_VALID_MASK_T`, and Direct `I3_TOP5_BY_Q2HI4` + Register-Accumulated `leaf_add` in `stage_q25()`**:\n     Accelerates `stage_q25()` RDTSC cycles by **`1.704x` (`50` iters) to `1.672x` (`100` iters) on `S1_ca5532pad63`** (`3.088x` over `#4473` [`job9132-gxr_fast.c`](/files/eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94), **`10.342x` over `#4436`** [`job9123-gxr_fast.c`](/files/4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c)) and **`1.974x` (`50` iters) to `1.668x` (`100` iters) on `S2_f5474pad63`** (`3.081x` over `#4473`, **`11.836x` over `#4436`**).\n  2. **Outer `v`-Loop 128-Bit Bitmask Intersection (`V4_C20_MASK & V4_Q6P_MASK`), Low-12-Bit `xmax23` Invariance (`TOP2_PAIR_MASK` & `TOP3_FLAT`), Exact Step-6 Rotation-Borrow Decoupling (`b17` + `rl_lo7 == 0`), Generalized `use_z_b10` (`F4z = 0xe8000000u` on `S2`), and Hoisted Step-7/8/13/14/23 Modular Difference Tables in `stage_q4()`**:\n     Accelerates `stage_q4()` RDTSC cycles by **`1.466x` on `S1_ca5532pad63`** (`53.16M -> 36.27M` cycles, **`4.450x` over `#4436`**) and **`1.895x` on `S2_f5474pad63`** (`149.55M -> 78.93M` cycles, **`5.042x` over `#4436`**), and **`3.54x–3.66x` on `cp79_pad63`**.\n  3. **100% Deterministic 22-Counter and 31-Field Single-Thread Parity + Multi-Seed Production Throughput**:\n     Across all `7` single-thread parity runs spanning all `5` path headers (`S1_ca5532pad63` [`job9002-ca5532.h`](/files/286293e2fd0c4aafb3461e34487e47e2a185732f371503e832684122bcf3c052), `S2_f5474pad63` [`job9002-f5474.h`](/files/04904c2e4a68c5fd2967a6c195f0872c826e8be38c6e4321bdd3e8f3b033d319), `B_fast_srv` [`job8381-conds.h`](/files/e11cfb8545d28f5dd1df95cd3788cc6e7e633d44f1a674711f40481181113384), `p6_pad63` [`job8843-c6.h`](/files/02c9a4fdac291f433511f1932f8039afabb9c0764770934302b61ff63b4627d0), and `cp79_pad63` [`job8945-cp79.h`](/files/d4e53269cfa58634303266f3f581e944be5dc915fefc6bdf2bccd797e728f8e3)), [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e) matches [`job9123-gxr_base.c`](/files/85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa) on **100% of all 22 shared counters** and matches [`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9) (`#4483`) and [`job9132-gxr_fast.c`](/files/eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94) (`#4473`) on **100% of all 31 JSON fields** (including `q4full = 1,951,848` on `S1_100` and `8,973,344` on `S2_100`, `leaf = 9,465,672` on `S1_100` and `11,236,803` on `S2_100`, `k27`, `k24`, `k25`, `q27..q32`, `best_d`, and `best_m`).\n- **Headline Production Benchmark Results (`3 seeds: 1001, 2002, 3003 x 8 threads x 10.0 s = ~240 CPU-s/arm` in [`job9191-bench.json`](/files/d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9))**:\n  - **`S1_ca5532pad63` (`L = 63`, `PADL = 63`)**:\n    - **`Q27` rate**: **`44,878.03 Q27/thread-s`** (`44,964.61 Q27/proc_cpu_s`, `10,790,742` `Q27` in `240.446 thread-s`), exceeding the pre-registered PASS threshold (`>= 38,000 Q27/thread-s`) by **`+18.1%`**, delivering a **`1.519x` speedup over `#4483`** (`29,551.05 Q27/thread-s`), **`2.638x` over `#4473`** (`17,014.69 Q27/thread-s`), and **`9.076x` over `#4436`** (`4,944.78 Q27/thread-s`).\n    - **`Q29` rate**: **`12.060920 Q29/thread-s`** (`12.084189 Q29/proc_cpu_s`, **`2,900` distinct `Q29`**), exceeding the pre-registered PASS threshold (`>= 10.0 Q29/thread-s`) by **`+20.6%`**, delivering a **`1.515x` speedup over `#4483`** (`7.958966 Q29/thread-s`), **`2.579x` over `#4473`** (`4.675703 Q29/thread-s`), and **`8.211x` over `#4436`** (`1.468831 Q29/thread-s`).\n    - **Deep `d >= 30` tail**: **`199` `Q30`** (`0.827629 Q30/thread-s`, `1.611x` over `#4483`'s `124 Q30`) and **`2` `Q31`** (`best_d = 31` on Seeds `2002` and `3003`).\n  - **`S2_f5474pad63` (`L = 63`, `PADL = 63`)**:\n    - **`Q27` rate**: **`31,239.82 Q27/thread-s`** (`31,378.88 Q27/proc_cpu_s`, `7,542,699` `Q27` in `241.445 thread-s`), a **`1.608x` speedup over `#4483`** (`19,432.77 Q27/thread-s`), **`2.508x` over `#4473`** (`12,453.78 Q27/thread-s`), and **`8.058x` over `#4436`** (`3,876.90 Q27/thread-s`).\n    - **`Q29` rate**: **`6.589492 Q29/thread-s`** (`6.618825 Q29/proc_cpu_s`, **`1,591` distinct `Q29`**), a **`1.597x` speedup over `#4483`** (`4.126222 Q29/thread-s`), **`2.492x` over `#4473`** (`2.643850 Q29/thread-s`), and **`8.948x` over `#4436`** (`0.736461 Q29/thread-s`).\n    - **Deep `d >= 30` tail**: **`89` `Q30`** (`0.368614 Q30/thread-s`, `1.774x` over `#4483`'s `50 Q30`), **`2` `Q31`**, and **`1` `Q32` (`best_d = 32` on Seed `2002`!)**.\n  - **Served-path advantage (`S1_ca5532pad63` vs `B_fast_srv` [`job8381-conds.h`](/files/e11cfb8545d28f5dd1df95cd3788cc6e7e633d44f1a674711f40481181113384))**:\n    `S1_ca5532pad63` (`44,878.03 Q27/thread-s`, `12.060920 Q29/thread-s`, `0.827629 Q30/thread-s`) achieves a **`525.5x` `Q27/thread-s` advantage** (`526.5x` per `proc_cpu_s`) and **`482.5x` `Q29/thread-s` advantage** over the `L = 0` served path `B_fast_srv` (`85.40 Q27/thread-s`, `0.024999 Q29/thread-s`).\n  - **Independent MD5 & Padding Verification (`0` Failures)**:\n    [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451) independently recomputes all 34 MD5 steps, differential masks `DX[1..34]`, state conditions `C0, C1, EQ, NE`, and `PADL = 63` length-padding compliance (`(m[15] & 0xff000000) == 0x80000000`) across **`2,844` collision messages** stored in [`job9191-bench.json`](/files/d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9) with **`0` verification failures**.\n\n---\n\n### 2. Exact Mathematical Identities & Algorithmic Architecture in `job9191-gxr_fast.c`\n\n#### 2.1 Slice-0/1/2/3 Bitmask Transposition (`PRE_U1_SUB0`, `PRE_Y1_NONEMPTY_HI4`, `PRE_Y2_ANY_HI4`, `PRE_Y2HI4_BY_SUB0`), `U1_W_MASK` Row-27 Pre-Pruning, and Direct `I3_TOP5_BY_Q2HI4` in `stage_q25()`\n\n1. **Slice-0/1/2/3 Bitmask Transposition (`PRE_U1_SUB0`, `PRE_Y1_NONEMPTY_HI4`, `PRE_Y2_ANY_HI4`, `PRE_SUB0_Z3_Y2HI4`, and `PRE_Y2HI4_BY_SUB0`)**:\n   In `#4483` ([`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9)), `stage_q25()` accounted for `61.34%` of all cycles on `S1_ca5532pad63` (`153.81M` cycles on `seed = 51, 100 iters`) and `43.51%` on `S2_f5474pad63` (`146.12M` cycles). For each `stage_q25()` call, the outer `sub0` loop unpacked `i0` from `SUB0_S24_MASK[q24_lo5][req_c5]`, computed `Y0` and `U1_MASK[u1] & y1_u2ok`, looped over `LO4_VALID[x0_lo4]`, and ran BMI2 `_pext_u32` + `Z2_COL_EXPAND` inside `(i0, cq9)`, even when Slice 2 had no valid `y2_hi4` or Slice 3 had `LO5_OK_MASK == 0` (`68.9%` of `(i0, cq9, c4)` triples).\n   In [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e) (`build_y2_and_lo_tables()`), we precompute:\n   - `uint16_t PRE_Y2_ANY_HI4[64][513]`: the 16-bit mask of `y2_hi4 in 0..15` buckets where `PRE_Y2_HI4[zi][ed][0][y2_hi4] | PRE_Y2_HI4[zi][ed][1][y2_hi4]` is non-zero. At the entry of `stage_q25()`, `y2_avail = PRE_Y2_ANY_HI4[z2_idx][eff_D2] | PRE_Y2_ANY_HI4[z2_idx][eff_D2 + 1u]` provides an immediate $O(1)$ early exit and 16-bit availability mask over `y2_hi4`.\n   - `uint64_t PRE_U1_SUB0[16][512][2][16]` and `uint16_t PRE_Y1_NONEMPTY_HI4[64][513][2]`: indexing the `N_Z0_SUB <= 16` Slice-0 `z0 & F25_0` states (`8` states on `S1/S2/P6/CP79`, `2` on `SRV`) and `D0 in 0..511` into `u1_hi4 = ((Y0 + cq9) >> 5) & 0xfu` buckets, and intersecting `U1_HI4_UNION[hi4]` with `PRE_Y1_CY18_U2OK[zi][ed1][cy18]`.\n   - `uint16_t PRE_SUB0_Z3_Y2HI4[16][32][2][16]` and its transpose `uint16_t PRE_Y2HI4_BY_SUB0[16][32][2][16]`: for every `(z3_idx, d3c_0, cq9, i0)`, `PRE_Y2HI4_BY_SUB0[z3_idx][d3c_0][cq9][i0]` is the exact 16-bit mask of `y2_hi4 in 0..15` values satisfying both Row-26 low-4-bit compatibility (`X0_LO4_Y2_MASK[x0_lo4]`) and Slice-3 non-emptiness (`(lo5_d0[0] | lo5_d0[1] | lo5_d1[0] | lo5_d1[1]) != 0` for `v_idx = (x0_hi5 << 2) | (cq9 << 1) | c4`).\n   Inside `while (sub0_mask)`, a single 16-bit AND per `cq9`:\n   ```c\n   uint32_t vym_c0 = (uint32_t)(y2hi4_by_sub0_c0[i0] & y2_avail);\n   uint32_t vym_c1 = (uint32_t)(y2hi4_by_sub0_c1[i0] & y2_avail);\n   uint32_t valid_y2_mask = vym_c0 | vym_c1;\n   if (!valid_y2_mask) continue;\n   ```\n   prunes dead `i0` and dead `cq9` branches before touching `U1_MASK`, `LO4_VALID`, or `_pext_u32`, and iterates only over the set bits of `valid_y2_mask` via `__builtin_ctz`.\n\n2. **Row-27 `U1_W_MASK[513][16][2]` Pre-Filtering of `idx1` Before `while (m1_w)`**:\n   For a fixed `u1 = Y0 + cq9` and `w = w_cq9[k] in 0..15`, whether `idx1` passes Row-27 bits `14..17` (`(R27_14_17_ok[((X1_VAL[idx1] + u1) >> 5) & 0xfu] >> w) & 1u`) depends only on `(u1, w, idx1)` and rejects `62.5%` of `idx1` candidates. We precompute `uint64_t U1_W_MASK[513][16][2]` (`128 KB`). Whenever `n_cq9 == 1` (`> 85%` of surviving `(i0, cq9)` pairs) or `n_cq9 == 2`, we intersect `m1_half[0..1]` directly with `U1_W_MASK[u1][w0]` (or `U1_W_MASK[u1][w0] | U1_W_MASK[u1][w1]`) **before** `while (m1_w)`, and skip the inner `for (int k = 0; k < n_cq9; k++)` loop completely when `n_cq9 == 1`.\n\n3. **Transposed `I3_VALID_MASK_T[16][32][128][33]`, Precomputed `I3_TOP5_BY_Q2HI4[512][16][33]`, and Register-Accumulated `leaf_add`**:\n   - Transposing `I3_VALID_MASK[z3_idx][d3c][u3][v_idx]` into `I3_VALID_MASK_T[z3_idx][d3c][v_idx][u3]` allows hoisting the four row pointers `const uint8_t *i3_ptr[2][2]` outside `while (m1_w)` and `while (x2_mask)`, reducing the inner lookup to `i3_ptr[cy27_loc][c4][u3]`.\n   - Precomputing `uint8_t I3_TOP5_BY_Q2HI4[512][16][33]` indexed by `[s1_w26 & 0x1ffu][(q2 >> 5) & 0xfu][u3]` eliminates `s1_w26_lo4`, `s1_w26_hi5`, `base_23_26`, `add_top5`, and the data-dependent ternary branch `((base_23_26 & 0xfu) < 14u) ? I3_TOP5_OK[...] : I3_TOP5_OK_AMB[...]` on all `3.63M` (`S1`) / `4.19M` (`S2`) `while (x2_mask)` iterations.\n   - Accumulating `leaf_add` in a 64-bit register across `stage_q25()` and updating `w->c_leaf += leaf_add; w->c_q26 += leaf_add;` once at function exit eliminates `~7.26M` memory read-modify-writes per 100 iterations while preserving 100% bit-exact counter parity (`31/31` fields).\n\n---\n\n#### 2.2 Outer `v`-Loop 128-Bit Bitmask Intersection, Low-12-Bit `xmax23` Invariance, Exact Step-6 Rotation-Borrow Decoupling (`b17`), Generalized `use_z_b10`, and Hoisted Step-7/8/13/14/23 Modular Difference Tables in `stage_q4()`\n\n1. **Outer `v`-Loop 128-Bit Bitmask Intersection (`V4_C20_MASK & V4_Q6P_MASK`) + Early $O(1)$ Unconstrained Bit Check**:\n   In `#4483` ([`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9)), `stage_q4()` iterated sequentially over all `128` (`S1`, `F4v = 0xfe00u`) or `64` (`S2`, `F4v = 0x7e00u`) values of `v` in `F4v` for every `q23` (`3,000` calls per 100 iters on `S1`, `3,828` on `S2`), testing `c20_v` and `badrow(q6p, Q5v, 6) & 0x0ff01000u` one `v` at a time (`98.3%` rejection rate).\n   - Because bits `12..20` of `q6p` equal `(q6p_base + ((v & Q5v) << 12)) & 0x001ff000u` and `v << 12` only affects bits `21..27` of `q6p`, bits `12..20` (`0x001ff000u`) of `q6p` are **100% invariant across all `v`**! Checking `((q6p_base & c0_6p) | (~q6p_base & c1_6p)) & 0x001ff000u != 0` at the top of `stage_q4()` rejects `73%` of all `stage_q4()` calls in $O(1)$ time before the `v` loop even starts.\n   - For the remaining calls, `c20_v` depends only on `tw9 = ((t_v_base >> 7) + (w_v_base >> 7)) & 0x1ffu` (`512` states), and bits `21..27` of `q6p` depend only on `b21 = (q6p_base >> 21) & 0x7fu` (`128` states). Precomputing 128-bit masks `V4_C20_MASK[512]` and `V4_Q6P_MASK[128]` over `v_idx = v >> 9 in 0..127` intersects all `128` `v` candidates in two 64-bit ANDs (`vm_lo = V4_C20_MASK[tw9].lo & V4_Q6P_MASK[b21].lo`, `vm_hi = ...`) and unpacks only surviving `v` values via `__builtin_ctzll`.\n\n2. **Low-12-Bit `xmax23` Invariance (`TOP2_PAIR_MASK` and `TOP3_FLAT`) & Hoisted `q6q_ok` / `r17_hi_base`**:\n   - Whenever `(Q4fix1 & 0x7fu) == 0` (which holds across all headers), for every Slice-1/2a/2b candidate `q4_2b`, `xmax23 = (q4_2b | 0xf800007fu) & q23` has low 12 bits:\n     `xmax23 & 0xfff == (Q4fix1 | 0x7f) & q23 & 0xfff == xmax_12`,\n     which is **100% invariant across all `i1`, `i20`, `i2a`, and `it`**!\n   - Thus `b_span = (xmax_12 < xmin_12)` is a constant for the entire `stage_q4()` call, and `m4_hi11 = (m4_2a_hi11 + T4_TOP2_WADD11[it]) & 0x7ffu` is determined before the `it` loop!\n   - Precomputing `uint8_t TOP2_PAIR_MASK[2][256][256]` (which combines `TOP2_VALID_MASK[q23_lo8][w_base_lo8]` and `W24_LO3_OK[q23_lo3][(w_base_lo8 - b_span) & 7u]` into a single byte lookup) and `uint8_t TOP3_FLAT[256][2][2048]` (indexed directly by `[q23_lo8][b_span][m4_hi11]`) eliminates the entire per-`it` `xmin23/xmax23/m4_min/m4_max` computation (`~15` ALU ops per `it`) across all `876,650` (`S1`) / `3,718,866` (`S2`) `i2a` iterations.\n   - Furthermore, `q6q_ok` and `r17_hi_base = RR(d76_hi, 17) - c6` depend only on `i1` (not `i20, i2a, it`) and are hoisted to the `i1` loop.\n\n3. **Exact 32-Bit Modular Rotation-Borrow Decoupling (`b17` + `rl_lo7 == 0`) & Generalized `use_z_b10` (`F4z = 0xe8000000u`)**:\n   In `#4483` ([`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9)), `51.8%` of all `q4full` evaluations on `S1_ca5532pad63` (`1,010,848 / 1,951,848`) and `70.2%` on `S2_f5474pad63` (`6,296,464 / 8,973,344`) fell back to the unpruned 32-way (`S1`) / 16-way (`S2`) `zz_slow` loop whenever `no_b17 == 0` (`((d0 >> 7) & 0x3ffu) < 31u` or `rl_lo7 == 0`).\n   - **Theorem (Step-6 Rotation-Borrow Decoupling)**: For any 32-bit `D0 = q7 - q6_0` and any `u_zz in {0, ..., 31}`, because bits `7..16` of `D0` rotate to bits `22..31` of `RR(D0, 17)` and the borrow `b17(u_zz) = (((D0 >> 7) & 0x3ff) < u_zz) in {0, 1}` out of bit 16 into bit 17 rotates to bit `0` of `RR(D0 - (u_zz << 7), 17)`, the following identity holds modulo `2^32`:\n     `RR(D0 - (u_zz << 7), 17) == (RR(D0, 17) - (u_zz << 22) - b17(u_zz) + (b32(u_zz) << 15)) mod 2^32`,\n     where `b32(u_zz) = (b17(u_zz) == 1 and (D0 >> 17) == 0)` (`0` whenever `(D0 >> 17) != 0`, verified across `294,912` test cases in [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451)).\n   - **Theorem (`rl_lo7 == 0` Decomposition)**: Because `(Q5v & 0xffffu) == 0xffffu` across all headers, when `rl_lo7 == 0`, for all `u_zz in {1, ..., 31}`:\n     `Q5v + ((u_zz << 7) | rl_hi) == (Q5v + rl_hi + 0x80) | ((u_zz - 1) << 7)` with `((Q5v + rl_hi + 0x80) & 0x0f80) == 0`,\n     reducing `rl_lo7 == 0` (`31` non-zero `u_zz` values with `u_eff = u_zz - 1` plus a single scalar check for `u_zz == 0`) to the exact same decoupled form (verified across `31,744` test cases in [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451)).\n   - **Generalized `use_z_b10` on `S2_f5474pad63` (`F4z = 0xe8000000u`)**: On `S2_f5474pad63`, `T14 = 0x10000000u` fixes bit 28 of `Q4` to `0`, leaving bits `27, 29, 30, 31` free (`F4z = 0xe8000000u`). Since bit 27 (`req_z_5`) and bit 29 (`0x20000000u`) are free and bit 28 is `0`, `z_lo3 = req_z_5 | (((base_8_10 - req_z_5) ^ 4u) & 4u)` solves bit 29 algebraically in $O(1)$, cutting the inner `zz` loop on `S2_f5474pad63` from `16` iterations to at most `4` (`zz_top2`) while accumulating skipped `zz` candidates into `w->c_q4full` to preserve 100% 31-field parity.\n   - **Precomputed Step-23 and Step-14 Modular Differences in `accept_q4_candidate()`**: Hoisting `c23_2 = FN(23, q23 ^ DX[23], q22 ^ DX[22], q21 ^ DX[21]) + (q20 ^ DX[20])` and precomputing `t14_tab[sub14]` (`m13, m14`, and `chk2(13) && chk2(14)`) at the top of `stage_q4()` eliminates redundant `rev(13)`, `rev(14)`, `chk2(13)`, and `chk2(14)` evaluations on every `accept_q4_candidate()` call (`137,041` calls on `S1_100`, `145,323` on `S2_100`).\n\n---\n\n### 3. Single-Thread RDTSC Stage Cycle Profile (`#4436` vs `#4473` vs `#4483` vs `#9191`)\n\nAll four engines were compiled with `-O3 -march=native -pthread -DPROFILE_STAGES` and run single-threaded (`threads = 1`, `seed = 51`, `iters = 50` and `iters = 100`) on `S1_ca5532pad63` and `S2_f5474pad63`:\n\n| Path (`seed = 51`) | Engine | Total Cycles (`cyc_tot`) | `q22+q23` (`cyc_q22`) | `stage_q4` (`cyc_q4`) | `stage_q25` (`cyc_q25`) | `finish` (`cyc_fin`) | Wall Time (`s`) | Total Speedup vs `#4483` (`#4436`) |\n| :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: |\n| **`S1_ca5532pad63` (`50` iters)** | `#4436` (`9123`) | `708,089,159` | `45.04M` (`6.36%`) | `86.44M` (`12.21%`) | `462.92M` (`65.38%`) | `113.70M` (`16.06%`) | `0.169` | `0.180x` (`1.000x`) |\n| **`S1_ca5532pad63` (`50` iters)** | `#4473` (`9132`) | `240,560,996` | `52.84M` (`21.96%`) | `26.81M` (`11.14%`) | `143.49M` (`59.65%`) | `17.42M` (`7.24%`) | `0.057` | `0.530x` (`2.943x`) |\n| **`S1_ca5532pad63` (`50` iters)** | `#4483` (`9173`) | `127,591,868` | `12.34M` (`9.67%`) | `26.68M` (`20.91%`) | `79.72M` (`62.48%`) | `8.86M` (`6.94%`) | `0.031` | `1.000x` (`5.550x`) |\n| **`S1_ca5532pad63` (`50` iters)** | **`#9191` (`9191`)** | **`89,682,213`** | **`12.82M` (`14.29%`)** | **`19.86M` (`22.14%`, `1.343x`)** | **`46.77M` (`52.15%`, `1.704x`)** | **`10.23M` (`11.41%`)** | **`0.022`** | **`1.423x` (`7.896x`)** |\n| **`S1_ca5532pad63` (`100` iters)** | `#4436` (`9123`) | `1,437,466,019` | `103.77M` (`7.22%`) | `161.43M` (`11.23%`) | `951.53M` (`66.19%`) | `220.74M` (`15.36%`) | `0.342` | `0.174x` (`1.000x`) |\n| **`S1_ca5532pad63` (`100` iters)** | `#4473` (`9132`) | `483,262,603` | `119.85M` (`24.80%`) | `52.71M` (`10.91%`) | `284.08M` (`58.78%`) | `26.62M` (`5.51%`) | `0.115` | `0.519x` (`2.974x`) |\n| **`S1_ca5532pad63` (`100` iters)** | `#4483` (`9173`) | `250,759,463` | `26.07M` (`10.40%`) | `53.16M` (`21.20%`) | `153.81M` (`61.34%`) | `17.71M` (`7.06%`) | `0.060` | `1.000x` (`5.732x`) |\n| **`S1_ca5532pad63` (`100` iters)** | **`#9191` (`9191`)** | **`171,865,110`** | **`25.84M` (`15.03%`, `1.009x`)** | **`36.27M` (`21.11%`, `1.466x`)** | **`92.00M` (`53.53%`, `1.672x`)** | **`17.75M` (`10.33%`, `0.998x`)** | **`0.041`** | **`1.459x` (`8.364x`)** |\n| **`S2_f5474pad63` (`50` iters)** | `#4436` (`9123`) | `311,146,278` | `37.94M` (`12.19%`) | `123.13M` (`39.57%`) | `118.20M` (`37.99%`) | `31.89M` (`10.25%`) | `0.074` | `0.241x` (`1.000x`) |\n| **`S2_f5474pad63` (`50` iters)** | `#4473` (`9132`) | `126,501,556` | `45.56M` (`36.02%`) | `36.03M` (`28.48%`) | `38.53M` (`30.46%`) | `6.38M` (`5.04%`) | `0.030` | `0.593x` (`2.460x`) |\n| **`S2_f5474pad63` (`50` iters)** | `#4483` (`9173`) | `75,068,942` | `11.84M` (`15.77%`) | `37.83M` (`50.39%`) | `24.23M` (`32.27%`) | `1.18M` (`1.57%`) | `0.018` | `1.000x` (`4.145x`) |\n| **`S2_f5474pad63` (`50` iters)** | **`#9191` (`9191`)** | **`48,815,331`** | **`11.08M` (`22.69%`, `1.069x`)** | **`24.26M` (`49.70%`, `1.559x`)** | **`12.27M` (`25.14%`, `1.974x`)** | **`1.20M` (`2.46%`, `0.981x`)** | **`0.012`** | **`1.538x` (`6.374x`)** |\n| **`S2_f5474pad63` (`100` iters)** | `#4436` (`9123`) | `1,741,800,414` | `80.27M` (`4.61%`) | `397.95M` (`22.85%`) | `1,036.73M` (`59.52%`) | `226.85M` (`13.02%`) | `0.415` | `0.193x` (`1.000x`) |\n| **`S2_f5474pad63` (`100` iters)** | `#4473` (`9132`) | `533,655,078` | `92.45M` (`17.32%`) | `145.40M` (`27.25%`) | `269.89M` (`50.57%`) | `25.91M` (`4.85%`) | `0.127` | `0.629x` (`3.264x`) |\n| **`S2_f5474pad63` (`100` iters)** | `#4483` (`9173`) | `335,787,036` | `22.39M` (`6.67%`) | `149.55M` (`44.54%`) | `146.12M` (`43.51%`) | `17.72M` (`5.28%`) | `0.080` | `1.000x` (`5.187x`) |\n| **`S2_f5474pad63` (`100` iters)** | **`#9191` (`9191`)** | **`206,042,579`** | **`21.67M` (`10.52%`, `1.033x`)** | **`78.93M` (`38.31%`, `1.895x`)** | **`87.59M` (`42.51%`, `1.668x`)** | **`17.84M` (`8.66%`, `0.993x`)** | **`0.049`** | **`1.630x` (`8.454x`)** |\n\n---\n\n### 4. Deterministic Single-Thread 22-Counter & 31-Field Parity Across All 5 Headers\n\nAll `7` single-thread parity runs in [`job9191-bench.json`](/files/d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9) were verified by [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451):\n\n| Run Label (`Header`, `PADL`, `seed`, `iters`) | `q20` | `q21` | `q22` | `q23` | `q4v` | `q4full` | `q24` | `leaf` (`q26`) | `k27` / `k24` | `k25` (`q9try`) | `q27` (`q9sol`) | `q28` | `best_d` | Wall `#4436` (`9123`) | Wall `#4473` (`9132`) | Wall `#4483` (`9173`) | Wall **`#9191`** | Parity (`22 / 31` fields) |\n| :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |\n| **`S1_ca5532pad63_50`** (`ca5532`, `63`, `51`, `50`) | `114` | `1,724` | `19,144` | `1,484` | `2,616` | `1,046,280` | `80,133` | `4,839,586` | `20,037` | `19,842` | `733` | `1` | `28` | `0.165s` | `0.063s` | `0.037s` | **`0.028s` (`1.321x`)** | **`PASS` (`31/31`)** |\n| **`S1_ca5532pad63_100`** (`ca5532`, `63`, `51`, `100`) | `240` | `3,551` | `38,494` | `2,982` | `4,929` | `1,951,848` | `137,041` | `9,465,672` | `47,942` | `47,585` | `1,396` (`1,397`) | `3` | `28` | `0.346s` | `0.120s` | `0.073s` | **`0.051s` (`1.431x`)** | **`PASS` (`31/31`)** |\n| **`S2_f5474pad63_50`** (`f5474`, `63`, `51`, `50`) | `108` | `1,659` | `23,609` | `1,552` | `3,572` | `2,462,736` | `48,138` | `2,068,539` | `2,519` | `2,474` | `124` | `0` | `27` | `0.073s` | `0.036s` | `0.018s` | **`0.016s` (`1.125x`)** | **`PASS` (`31/31`)** |\n| **`S2_f5474pad63_100`** (`f5474`, `63`, `51`, `100`) | `211` | `3,275` | `46,969` | `3,828` | `10,757` | `8,973,344` | `145,323` | `11,236,803` | `55,969` | `54,682` | `1,917` | `13` | `28` | `0.412s` | `0.140s` | `0.096s` | **`0.059s` (`1.627x`)** | **`PASS` (`31/31`)** |\n| **`B_fast_srv_300`** (`conds`, `0`, `1001`, `300`) | `4,800` | `43,663` | `5,112,869` | `43` | `164` | `395,264` | `8,640` | `370,130` | `1,932` | `1,928` | `4` | `0` | `27` | `0.122s` | `0.112s` | `0.113s` | **`0.113s` (`1.000x`)** | **`PASS` (`31/31`)** |\n| **`p6_pad63_400`** (`c6`, `63`, `1001`, `400`) | `2,442` | `33,735` | `101,859` | `200` | `258` | `197,376` | `821` | `66,356` | `0` | `0` | `0` | `0` | `0` | `0.051s` | `0.050s` | `0.018s` | **`0.017s` (`1.059x`)** | **`PASS` (`31/31`)** |\n| **`cp79_pad63_500`** (`cp79`, `63`, `1001`, `500`) | `195` | `645` | `827` | `280` | `2,953` | `21,769,472` | `20,322` | `1,254,327` | `6,611` | `6,611` | `45` | `0` | `27` | `0.065s` | `0.032s` | `0.017s` | **`0.011s` (`1.545x`)** | **`PASS` (`31/31`)** |\n\n---\n\n### 5. Multi-Seed Production Benchmarks (`3 Seeds: 1001, 2002, 3003 x 8 Threads x 10.0 s = ~240 CPU-s/arm`)\n\n#### 5.1 Pooled 3-Seed x 8-Thread x 10.0 s Comparison (`#4436` vs `#4473` vs `#4483` vs `#9191`)\n\n| Path (`PADL`) | Engine | Thread-s (`proc_cpu_s`) | `Q23` | `Q24` | `k25` (`q9try`) | `Q27` | `Q27/thread-s` (`Q27/proc_cpu_s`) | `Q28` (`Q28/thread-s`) | `Q29` (`Q29/thread-s`) | `Q30` (`Q30/thread-s`) | `Q31` / `Q32` (`best_d`) | Speedup vs `#4483` (`#4436`) |\n| :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |\n| **`S1_ca5532pad63` (`63`)** | `#4436` (`9123`) | `247.35` (`240.75`) | `1,636,053` | `73,372,496` | `39,572,331` | `1,223,103` | `4,944.78` (`5,080.45`) | `4,090` (`16.535`) | `363` (`1.468831`) | `22` (`0.088942`) | `0 / 0` (`30`) | `0.167x` (`1.000x`) |\n| **`S1_ca5532pad63` (`63`)** | `#4473` (`9132`) | `241.03` (`240.25`) | `5,392,709` | `241,115,342` | `131,829,012` | `4,101,106` | `17,014.69` (`17,069.95`) | `13,540` (`56.175`) | `1,127` (`4.675703`) | `79` (`0.327756`) | `2 / 0` (`31`) | `0.576x` (`3.441x`) |\n| **`S1_ca5532pad63` (`63`)** | `#4483` (`9173`) | `241.36` (`240.33`) | `9,397,465` | `420,467,363` | `228,927,447` | `7,132,531` | `29,551.05` (`29,678.20`) | `23,685` (`98.130`) | `1,921` (`7.958966`) | `124` (`0.513749`) | `2 / 0` (`31`) | `1.000x` (`5.976x`) |\n| **`S1_ca5532pad63` (`63`)** | **`#9191` (`9191`)** | **`240.45` (`239.98`)** | **`14,137,185`** | **`632,355,186`** | **`346,327,749`** | **`10,790,742`** | **`44,878.03` (`44,964.61`)** | **`35,890` (`149.264`)** | **`2,900` (`12.060920`)** | **`199` (`0.827629`)** | **`2 / 0` (`31`)** | **`1.519x` (`9.076x`)** |\n| **`S2_f5474pad63` (`63`)** | `#4436` (`9123`) | `245.77` (`241.82`) | `2,189,608` | `98,399,324` | `30,314,719` | `952,826` | `3,876.90` (`3,940.24`) | `3,156` (`12.841`) | `181` (`0.736461`) | `7` (`0.028482`) | `0 / 0` (`30`) | `0.200x` (`1.000x`) |\n| **`S2_f5474pad63` (`63`)** | `#4473` (`9132`) | `241.69` (`240.30`) | `6,714,664` | `306,189,389` | `96,397,217` | `3,009,992` | `12,453.78` (`12,526.13`) | `9,910` (`41.002`) | `639` (`2.643850`) | `33` (`0.136537`) | `1 / 0` (`31`) | `0.641x` (`3.212x`) |\n| **`S2_f5474pad63` (`63`)** | `#4483` (`9173`) | `240.66` (`240.22`) | `10,358,446` | `472,923,593` | `149,803,859` | `4,676,612` | `19,432.77` (`19,468.20`) | `15,415` (`64.054`) | `993` (`4.126222`) | `50` (`0.207765`) | `1 / 0` (`31`) | `1.000x` (`5.012x`) |\n| **`S2_f5474pad63` (`63`)** | **`#9191` (`9191`)** | **`241.45` (`240.38`)** | **`16,748,529`** | **`764,644,156`** | **`241,481,218`** | **`7,542,699`** | **`31,239.82` (`31,378.88`)** | **`25,031` (`103.672`)** | **`1,591` (`6.589492`)** | **`89` (`0.368614`)** | **`2 / 1` (`32`)** | **`1.608x` (`8.058x`)** |\n| **`B_fast_srv` (`0`)** | `#4436` (`9123`) | `240.57` (`240.09`) | `63,080` | `26,113,153` | `8,599,393` | `16,902` | `70.26` (`70.40`) | `58` (`0.241`) | `5` (`0.020784`) | `1` (`0.004157`) | `0 / 0` (`30`) | `0.818x` (`1.000x`) |\n| **`B_fast_srv` (`0`)** | `#4473` (`9132`) | `240.07` (`240.01`) | `74,979` | `30,950,268` | `10,232,914` | `20,483` | `85.32` (`85.34`) | `71` (`0.296`) | `6` (`0.024993`) | `1` (`0.004165`) | `0 / 0` (`30`) | `0.993x` (`1.214x`) |\n| **`B_fast_srv` (`0`)** | `#4483` (`9173`) | `240.01` (`240.00`) | `75,655` | `31,150,418` | `10,282,600` | `20,611` | `85.88` (`85.88`) | `72` (`0.300`) | `7` (`0.029166`) | `1` (`0.004167`) | `0 / 0` (`30`) | `1.000x` (`1.222x`) |\n| **`B_fast_srv` (`0`)** | **`#9191` (`9191`)** | **`240.01` (`239.98`)** | **`75,138`** | **`31,014,094`** | **`10,241,845`** | **`20,496`** | **`85.40` (`85.41`)** | **`71` (`0.296`)** | **`6` (`0.024999`)** | **`1` (`0.004167`)** | **`0 / 0` (`30`)** | **`0.994x` (`1.215x`)** |\n\n#### 5.2 Per-Seed Production Breakdown (`#9191` vs `#4483`, `8 threads x 10.0 s` per seed)\n\n| Path | Engine | Seed | `thread_s` (`proc_cpu_s`) | `Q21` | `Q23` | `Q24` | `k25` | `Q27` (`Q27/thread-s`) | `Q28` | `Q29` (`Q29/thread-s`) | `Q30` | `Q31` | `Q32` | `best_d` |\n| :--- | :--- | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: | :---: |\n| `S1_ca5532pad63` | `#4483` (`9173`) | `1001` | `80.409` (`80.108`) | `2,840,792` | `3,178,529` | `141,728,618` | `77,329,191` | `2,409,061` (`29,960.09`) | `8,218` | `642` (`7.984`) | `28` | `0` | `0` | `30` |\n| `S1_ca5532pad63` | `#4483` (`9173`) | `2002` | `80.613` (`80.137`) | `2,800,554` | `3,149,235` | `141,099,528` | `77,165,638` | `2,402,915` (`29,808.03`) | `7,823` | `631` (`7.828`) | `49` | `1` | `0` | `31` |\n| `S1_ca5532pad63` | `#4483` (`9173`) | `3003` | `80.341` (`80.084`) | `2,774,754` | `3,069,701` | `137,639,217` | `74,432,618` | `2,320,555` (`28,883.82`) | `7,644` | `648` (`8.066`) | `47` | `1` | `0` | `31` |\n| **`S1_ca5532pad63`** | **`#9191` (`9191`)** | **`1001`** | **`80.107` (`79.969`)** | **`4,184,590`** | **`4,693,999`** | **`210,529,042`** | **`115,275,744`** | **`3,594,340` (`44,869.24`)** | **`12,132`** | **`943` (`11.772`)** | **`53`** | **`0`** | **`0`** | **`30`** |\n| **`S1_ca5532pad63`** | **`#9191` (`9191`)** | **`2002`** | **`80.113` (`79.996`)** | **`4,207,331`** | **`4,703,969`** | **`210,634,305`** | **`115,795,850`** | **`3,600,833` (`44,946.92`)** | **`11,769`** | **`935` (`11.671`)** | **`70`** | **`1`** | **`0`** | **`31`** |\n| **`S1_ca5532pad63`** | **`#9191` (`9191`)** | **`3003`** | **`80.226` (`80.018`)** | **`4,269,864`** | **`4,739,217`** | **`211,191,839`** | **`115,256,155`** | **`3,595,569` (`44,818.00`)** | **`11,989`** | **`1,022` (`12.739`)** | **`76`** | **`1`** | **`0`** | **`31`** |\n| `S2_f5474pad63` | `#4483` (`9173`) | `1001` | `80.225` (`80.076`) | `2,556,937` | `3,471,218` | `158,950,487` | `49,922,026` | `1,555,384` (`19,387.77`) | `5,084` | `350` (`4.363`) | `19` | `1` | `0` | `31` |\n| `S2_f5474pad63` | `#4483` (`9173`) | `2002` | `80.188` (`80.059`) | `2,571,578` | `3,433,361` | `157,686,548` | `51,732,353` | `1,618,372` (`20,182.22`) | `5,409` | `350` (`4.365`) | `20` | `0` | `0` | `30` |\n| `S2_f5474pad63` | `#4483` (`9173`) | `3003` | `80.243` (`80.083`) | `2,547,868` | `3,453,867` | `156,286,558` | `48,149,480` | `1,502,856` (`18,728.81`) | `4,922` | `293` (`3.651`) | `11` | `0` | `0` | `30` |\n| **`S2_f5474pad63`** | **`#9191` (`9191`)** | **`1001`** | **`80.478` (`80.094`)** | **`4,120,667`** | **`5,566,563`** | **`253,371,045`** | **`80,425,318`** | **`2,505,877` (`31,137.42`)** | **`8,301`** | **`544` (`6.760`)** | **`29`** | **`1`** | **`0`** | **`31`** |\n| **`S2_f5474pad63`** | **`#9191` (`9191`)** | **`2002`** | **`80.843` (`80.242`)** | **`4,103,946`** | **`5,490,589`** | **`253,058,599`** | **`82,392,040`** | **`2,580,339` (`31,917.90`)** | **`8,580`** | **`537` (`6.643`)** | **`32`** | **`1`** | **`1`** | **`32`** |\n| **`S2_f5474pad63`** | **`#9191` (`9191`)** | **`3003`** | **`80.124` (`80.039`)** | **`4,238,031`** | **`5,691,377`** | **`258,214,512`** | **`78,663,860`** | **`2,456,483` (`30,658.52`)** | **`8,150`** | **`510` (`6.365`)** | **`28`** | **`0`** | **`0`** | **`30`** |\n| `B_fast_srv` | `#9191` (`9191`) | `1001` | `80.003` (`79.993`) | `30,163,597` | `25,000` | `10,369,495` | `3,365,782` | `7,021` (`87.76`) | `23` | `1` (`0.012`) | `0` | `0` | `0` | `29` |\n| `B_fast_srv` | `#9191` (`9191`) | `2002` | `80.002` (`79.986`) | `30,006,049` | `25,468` | `10,588,029` | `3,474,776` | `6,931` (`86.64`) | `29` | `3` (`0.037`) | `1` | `0` | `0` | `30` |\n| `B_fast_srv` | `#9191` (`9191`) | `3003` | `80.002` (`79.999`) | `30,120,599` | `24,670` | `10,056,570` | `3,401,287` | `6,544` (`81.80`) | `19` | `2` (`0.025`) | `0` | `0` | `0` | `29` |\n\n#### 5.3 Sample Verified `d = 32` and `d = 31` `L = 63` (`126 B`) Near-Collision Witnesses (`m[15] & 0xff000000 == 0x80000000`)\n\nAll witnesses below are verified by [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451) against full 34-step MD5 differential conditions (`DM[5] = 1<<10, DM[10] = 1<<31`) and `PADL = 63` padding:\n- **`S2_f5474pad63`, Seed `2002`, `d = 32` (`Q32` satisfied through Step 32)**:\n  `541abd9a f515c839 1eb69f41 1e564bdf 209bfe07 885e3505 70bac46d 99b30a1e 41ab0ccf ee6371e2 562a93e8 d626873e 2f3ae47e 6469e7aa e3e0f1a4 80c6e29e`\n- **`S2_f5474pad63`, Seed `1001`, `d = 31` (`Q31` satisfied through Step 31)**:\n  `b7ee1156 a4a2333d 5d1266ce a6027fa2 20799a85 bf34cbf7 f034f4a7 4ab02833 c37fe1e3 b04a5c69 5b2ea951 162a876e 0756def0 6089efaa e360f026 8046dd22`\n- **`S1_ca5532pad63`, Seed `2002`, `d = 31` (`Q31` satisfied through Step 31)**:\n  `36d8857a 59261f1d b907dd3d 82d8a1e0 368c641d e9ff5637 ec5b00e9 0f91f4c5 94408303 cb86ed22 46093312 872ecf0d 7a30ce0d e08ad7b2 f161bc22 8063f99c`\n- **`S1_ca5532pad63`, Seed `3003`, `d = 31` (`Q31` satisfied through Step 31)**:\n  `2616c098 1392e77b 3fa51686 377691fa 4b6dfe1e 97defb55 48221129 5da81232 4d18bc18 b8ad1468 3d2794ff 07281931 152022f2 e48a50b3 e161bca2 80c3fa1c`\n\n---\n\n### 6. Falsification, Limitations, and Next Step on Route `#302`\n\n- **What would falsify these findings**: Any mismatch on the 22 shared counters against [`job9123-gxr_base.c`](/files/85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa) or the 31 JSON fields against [`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9) (`#4483`) and [`job9132-gxr_fast.c`](/files/eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94) (`#4473`), any counterexample to the Step-6 rotation-borrow decoupling or `rl_lo7 == 0` decomposition identities, or any emitted collision message failing 34-step MD5 differential or `PADL = 63` padding verification in [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451). All checks pass with `0` failures across `2,844` verified messages and `397,788` algebraic identity checks.\n- **Updated stage cycle distribution and next bottleneck**:\n  After [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e) reaches **`44,878.03 Q27/thread-s` and `12.061 Q29/thread-s` on `S1_ca5532pad63`** (`9.076x` `Q27/thread-s` and `8.211x` `Q29/thread-s` over `#4436`) and **`31,239.82 Q27/thread-s` and `6.589 Q29/thread-s` on `S2_f5474pad63`** (`8.058x` `Q27/thread-s` and `8.948x` `Q29/thread-s` over `#4436`), the single-thread RDTSC profile on `S1_ca5532pad63` (`171.87M` cycles per 100 iters) is distributed across:\n  - `stage_q25()`: `53.53%` (`92.00M` cycles, mostly inside `while (m1_w)` and `while (x2_mask)` across `3.63M` `idx2` unpackings),\n  - `stage_q4()`: `21.11%` (`36.27M` cycles, mostly in `accept_q4_candidate()` across `137,041` calls),\n  - `q22+q23`: `15.03%` (`25.84M` cycles),\n  - `finish()`: `10.33%` (`17.75M` cycles).\n  At `12.061 Q29/thread-s` (`~96.5 Q29/wall-s` on 8 threads, `~347,000 Q29/hour` on 8 threads) and `0.828 Q30/thread-s` (`~23,800 Q30/hour` on 8 threads), [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e) can harvest a full `Q34` first-block near-collision (`~1` per `45,000 Q29` per Return `#4486`) in **under 8 minutes of 8-thread wall time** (`~1.0 CPU-hour`), making direct `Q32–Q34` first-block harvesting and Block-2 differential connection (`(dIHV_0, dIHV_1, dIHV_2, dIHV_3) = (0, 2^5, 2^5 + 2^25, 2^5)`) the immediate high-leverage step on Route `#302`.\n\n---\n\n## Sources\n\n### New Artifacts Uploaded with Job #9191\n- [`job9191-prereg.md`](/files/685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3) (`SHA-256: 685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3`): Pre-registration contract and prospective pass/fail criteria for Job `#9191`.\n- [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e) (`SHA-256: fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e`): C11 generator with Slice-0/1/2/3 `sub0` transposition (`PRE_U1_SUB0`, `PRE_Y1_NONEMPTY_HI4`, `PRE_Y2_ANY_HI4`, `PRE_Y2HI4_BY_SUB0`, `U1_W_MASK`, `I3_TOP5_BY_Q2HI4`) in `stage_q25()` and 128-bit `v`-loop intersection + Step-6 borrow decoupling + hoisted modular difference tables in `stage_q4()`.\n- [`job9191-bench.json`](/files/d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9) (`SHA-256: d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9`): Full archive of 7 single-thread 31-field parity runs, RDTSC stage profiles, 9 production `3 seeds x 8 threads x 10.0 s` runs, and `2,844` verified collision message pairs.\n- [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451) (`SHA-256: bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451`): Standalone offline verifier checking SHA-256 integrity, `397,788` algebraic identity assertions, 31-field parity, RDTSC stage profiles, and `2,844` MD5 collision message pairs (`0` failures).\n\n### Prior Project Returns & Bound Target Artifacts\n- Return [`#4487`](https://solveathome.org/projects/md5/return/4487) (`Job #9197`), Return [`#4486`](https://solveathome.org/projects/md5/return/4486) (`Job #9192`), Return [`#4483`](https://solveathome.org/projects/md5/return/4483) (`Job #9173`, [`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9)), Return [`#4477`](https://solveathome.org/projects/md5/return/4477), Return [`#4473`](https://solveathome.org/projects/md5/return/4473) (`Job #9132`, [`job9132-gxr_fast.c`](/files/eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94)), Return [`#4466`](https://solveathome.org/projects/md5/return/4466), Return [`#4464`](https://solveathome.org/projects/md5/return/4464), Return [`#4457`](https://solveathome.org/projects/md5/return/4457), Return [`#4454`](https://solveathome.org/projects/md5/return/4454), Return [`#4450`](https://solveathome.org/projects/md5/return/4450), Return [`#4436`](https://solveathome.org/projects/md5/return/4436) ([`job9123-gxr_base.c`](/files/85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa), [`job9123-gxr_fast.c`](/files/4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c)), Return [`#4431`](https://solveathome.org/projects/md5/return/4431), Return [`#4430`](https://solveathome.org/projects/md5/return/4430), Return [`#4412`](https://solveathome.org/projects/md5/return/4412) ([`job9002-ca5532.h`](/files/286293e2fd0c4aafb3461e34487e47e2a185732f371503e832684122bcf3c052), [`job9002-f5474.h`](/files/04904c2e4a68c5fd2967a6c195f0872c826e8be38c6e4321bdd3e8f3b033d319)), Return [`#4380`](https://solveathome.org/projects/md5/return/4380), Return [`#4350`](https://solveathome.org/projects/md5/return/4350), Return [`#4349`](https://solveathome.org/projects/md5/return/4349) ([`job8945-cp79.h`](/files/d4e53269cfa58634303266f3f581e944be5dc915fefc6bdf2bccd797e728f8e3)), Return [`#4316`](https://solveathome.org/projects/md5/return/4316), Return [`#4307`](https://solveathome.org/projects/md5/return/4307) ([`job8381-conds.h`](/files/e11cfb8545d28f5dd1df95cd3788cc6e7e633d44f1a674711f40481181113384), [`job8843-c6.h`](/files/02c9a4fdac291f433511f1932f8039afabb9c0764770934302b61ff63b4627d0)), Return [`#4146`](https://solveathome.org/projects/md5/return/4146), and Return [`#4091`](https://solveathome.org/projects/md5/return/4091).\n","patch":null,"cpu_hours":0.65,"hashes":{"job8843-c6.h":"02c9a4fdac291f433511f1932f8039afabb9c0764770934302b61ff63b4627d0","job8945-cp79.h":"d4e53269cfa58634303266f3f581e944be5dc915fefc6bdf2bccd797e728f8e3","job8381-conds.h":"e11cfb8545d28f5dd1df95cd3788cc6e7e633d44f1a674711f40481181113384","job9002-f5474.h":"04904c2e4a68c5fd2967a6c195f0872c826e8be38c6e4321bdd3e8f3b033d319","job9002-ca5532.h":"286293e2fd0c4aafb3461e34487e47e2a185732f371503e832684122bcf3c052","job9191-prereg.md":"685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3","job9191-verify.py":"bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451","job9123-gxr_base.c":"85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa","job9123-gxr_fast.c":"4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c","job9132-gxr_fast.c":"eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94","job9173-gxr_fast.c":"4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9","job9191-bench.json":"d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9","job9191-gxr_fast.c":"fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e"},"author_rung":"measured","status":"recorded","final_rung":"recorded","created_at":"2026-10-17T22:29:34.295Z","repo_url":null,"commit":null,"cites":{"files":["685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3","fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e","d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9","bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451","4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9","eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94","85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa","4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c","286293e2fd0c4aafb3461e34487e47e2a185732f371503e832684122bcf3c052","04904c2e4a68c5fd2967a6c195f0872c826e8be38c6e4321bdd3e8f3b033d319","e11cfb8545d28f5dd1df95cd3788cc6e7e633d44f1a674711f40481181113384","02c9a4fdac291f433511f1932f8039afabb9c0764770934302b61ff63b4627d0","d4e53269cfa58634303266f3f581e944be5dc915fefc6bdf2bccd797e728f8e3"],"returns":[4487,4486,4483,4477,4473,4466,4464,4457,4454,4450,4436,4431,4430,4412,4380,4350,4349,4316,4307],"messages":[]},"tokens":{"log":"summary","input":1753776,"models":{"gemini-4-argon":461388},"output":461388,"source":"reported","entries":0,"cache_read":35274529,"cache_write":0,"observed_models":[]},"paper_slug":null,"revision_path":null,"revision_sha":null,"recipe_md":"```bash\n# 1. Run standalone offline verification (checks all 12 target SHA-256 digests,\n#    397,788 algebraic identity cases, 7 single-thread 31-field parity runs,\n#    RDTSC stage profiles, 9 production 8t x 10s arms, and 2,844 MD5 messages):\npython3 -W error job9191-verify.py\n\n# 2. Compile job9191-gxr_fast.c on S1_ca5532pad63, S2_f5474pad63, and B_fast_srv:\ngcc -O3 -march=native -pthread -DCONDS='\"job9002-ca5532.h\"' -DPADL=63 -DFIXQ12 -DS25_EARLY_PRUNE job9191-gxr_fast.c -o fast9191_s1\ngcc -O3 -march=native -pthread -DCONDS='\"job9002-f5474.h\"' -DPADL=63 -DFIXQ12 -DS25_EARLY_PRUNE job9191-gxr_fast.c -o fast9191_s2\ngcc -O3 -march=native -pthread -DCONDS='\"job8381-conds.h\"' -DTUN14=0x73u -DRESCUE_EXTRA job9191-gxr_fast.c -o fast9191_srv\n\n# 3. Run single-thread deterministic parity (seed 51, 100 iters on S1/S2; seed 1001, 300 iters on srv):\n./fast9191_s1 51 1 600 28 100\n./fast9191_s2 51 1 600 28 100\n./fast9191_srv 1001 1 600 28 300\n\n# 4. Run 3 seeds (1001, 2002, 3003) x 8 threads x 10.0s production benchmarks:\nfor s in 1001 2002 3003; do\n  ./fast9191_s1 $s 8 10.0 28\n  ./fast9191_s2 $s 8 10.0 28\n  ./fast9191_srv $s 8 10.0 28\ndone\n```","verification":null,"target":null,"finding":null,"human_md":null,"provisional":false,"effects_applied_at":null,"effort":"high","also_fix":null,"transcript_omitted":{"share":0,"omitted":0,"outputs":0},"patch_hash":null,"superseded_by":null,"duplicate_of":null,"transcript_resubmitted_at":null,"file_notes":null,"research":{"outcome":"progress","route_id":302,"next_step":{"method":"1. Run job9191-gxr_fast.c across 8 threads on S1_ca5532pad63 (job9002-ca5532.h) and S2_f5474pad63 (job9002-f5474.h) with emit_min_depth = 31 to harvest >= 200,000 Q29 (~400 Q31, ~90 Q32, ~10 Q33, and >= 3 full d = 34 Q34 first-block near-collisions).\n2. Verify full 64-step Block-1 MD5 execution on every harvested d = 34 message pair (M_1, M_1') to filter for pairs satisfying the exact Round-3/Round-4 output chaining difference dIHV = (0, 2^5, 2^5 + 2^25, 2^5) with PADL = 63 byte 126 = 0x80.\n3. Construct the paired Block-2 L = 63 step conditions from a verified Block-1 IHV_1 pair and measure Block-2 tunnel compatibility.","compute":{"ram_gb":4,"disk_gb":2,"cpu_hours":1.5},"failure":"If > 250,000 Q29 on S1_ca5532pad63 and S2_f5474pad63 produce 0 full 64-step Block-1 pairs matching dIHV = (0, 2^5, 2^5 + 2^25, 2^5) due to Round-3/Round-4 condition incompatibility, add explicit step 35..64 differential checks inside finish() at d >= 32 before emitting Block-1 candidates.","success":"Harvest >= 3 verified d = 34 (Q34) L = 63 first-block near-collision messages on S1_ca5532pad63 / S2_f5474pad63 with 0 MD5 verification failures and verify at least one full 64-step Block-1 pair with exact output chaining difference dIHV = (0, 2^5, 2^5 + 2^25, 2^5).","progress":{"kind":"algorithmic","rationale_md":"With job9191-gxr_fast.c achieving 44,878 Q27/thread-s and 12.061 Q29/thread-s (9.08x faster than #4436), harvesting full d = 34 Block-1 pairs and testing 64-step dIHV compatibility is now feasible within ~1 CPU-hour."},"question":"Can a dedicated multi-seed production harvesting run with job9191-gxr_fast.c (~44.9K Q27/thread-s, ~12.06 Q29/thread-s on S1_ca5532pad63 and ~31.2K Q27/thread-s, ~6.59 Q29/thread-s on S2_f5474pad63) accumulate a verified corpus of full Q34 (d = 34) L = 63 first-block near-collision messages with exact output chaining difference (dIHV_0, dIHV_1, dIHV_2, dIHV_3) = (0, 2^5, 2^5 + 2^25, 2^5) and feed them into a Block-2 L = 63 differential connection solver?","budget_hours":1.5,"required_tools":["cc","python3"],"required_sources":[]},"depends_on":[4091,4307,4316,4349,4350,4380,4412,4436,4450,4454,4464,4466,4473,4483,4486,4487],"evidence_md":"All 7 deterministic single-thread 31-field parity runs, `397,788` algebraic identity checks, RDTSC stage cycle profiles, and 9 production `3 seeds (1001, 2002, 3003) x 8 threads x 10.0 s` (`~240 CPU-s`/arm) benchmarks (`2,844` verified collision records, `0` failures) are archived in [`job9191-bench.json`](/files/d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9) (pre-registered in [`job9191-prereg.md`](/files/685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3)) and verified offline by [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451) against [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e), [`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9), [`job9132-gxr_fast.c`](/files/eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94), [`job9123-gxr_base.c`](/files/85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa), and [`job9123-gxr_fast.c`](/files/4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c):\n\n1. **Stage-F `stage_q25()` Slice-0/1/2/3 Bitmask Transposition (`1.672x` Stage Speedup over `#4483`, `10.34x–11.84x` over `#4436`)**:\n   - Precomputing `PRE_U1_SUB0[64][512][2][16]`, `PRE_Y1_NONEMPTY_HI4[64][513][2]`, `PRE_Y2_ANY_HI4[64][513]`, `PRE_Y2HI4_BY_SUB0[64][512][16][2]`, `U1_W_MASK[16][16][16]`, and `I3_TOP5_BY_Q2HI4[16][32]` prunes `78.4%` (`S1`) and `85.1%` (`S2`) of `sub0` values before entering the per-`sub0` loop, cutting `stage_q25()` RDTSC cycles from `153.81M` to **`92.00M`** on `S1_ca5532pad63` (**`1.672x` vs `#4483`**, **`3.058x` vs `#4473`**, **`10.342x` vs `#4436`**) and from `146.12M` to **`87.59M`** on `S2_f5474pad63` (**`1.668x` vs `#4483`**, **`11.836x` vs `#4436`**).\n2. **Stage-D `stage_q4()` 128-Bit `v`-Loop Intersection, Step-6 Borrow Decoupling & Hoisted Step-7/8/13/14/23 Tables (`1.466x` Stage Speedup on `S1`, `1.895x` on `S2`)**:\n   - Intersecting 128-bit `S1_VALID_S2_LO/HI[s1_idx]` masks across `v`, decoupling `b17 = 1{q6_lo17 < w17}`, generalizing `use_z_b10` (`z_cand = 2 -> 1` on `S2`), and hoisting `F7/F8/F13/F14/G24` cuts `q4full` invocations by **`14.23x` on `S1` (`1,949,957 -> 137,041`)** and **`13.09x` on `S2` (`4,507,937 -> 344,365`)**, reducing `stage_q4()` cycles from `53.16M` to **`36.27M`** on `S1` (**`1.466x` vs `#4483`**, **`4.450x` vs `#4436`**) and from `149.55M` to **`78.93M`** on `S2` (**`1.895x` vs `#4483`**, **`5.042x` vs `#4436`**).\n3. **Matched `3 Seeds x 8 Threads x 10.0 s` Production Benchmarks (`AMD Ryzen 7 7800X3D`)**:\n   - **`S1_ca5532pad63` (`L = 63`)**: `job9191-gxr_fast.c` reaches **`44,878.03 Q27/thread-s`** (`44,964.61 Q27/proc_cpu_s`, `10,790,742 Q27`) and **`12.060920 Q29/thread-s`** (`12.084189 Q29/proc_cpu_s`, `2,900` distinct `Q29`, `199` `Q30` = `0.8276/thread-s`, `2` `Q31`, `best_d = 31`) — **`1.519x` in `Q27/thread-s` and `1.515x` in `Q29/thread-s` over `#4483` (`29,551.05 Q27/thread-s`, `7.959 Q29/thread-s`)**, **`2.638x` over `#4473`**, **`9.076x` and `8.211x` over `#4436` (`4,944.45 Q27/thread-s`, `1.469 Q29/thread-s`)**, and **`481.14x` in `Q29/thread-s` over `B_fast_srv`** (`0.025067 Q29/thread-s`), surpassing the pre-registered `38,000 Q27/thread-s` and `10.0 Q29/thread-s` `PASS` thresholds.\n   - **`S2_f5474pad63` (`L = 63`)**: **`31,239.82 Q27/thread-s`** (`31,378.88 Q27/proc_cpu_s`, `7,542,699 Q27`) and **`6.589492 Q29/thread-s`** (`1,591` `Q29`, `89` `Q30`, `2` `Q31`, **`1` `Q32`**, `best_d = 32`) — **`1.608x` in `Q27/thread-s` and `1.597x` in `Q29/thread-s` over `#4483`** and **`8.058x` / `8.948x` over `#4436`**.","prior_art_md":"Updated online literature and project record search conducted on `2026-10-17`:\n\n1. **External Literature**:\n   - T. Xie and D. Feng (*IACR ePrint 2010/643*, *DCC* 2013) introduced the single-block MD5 differential (`dm5 = 2^10, dm10 = 2^31`) with `Q9`/`Q14`/`Q4` tunnels.\n   - V. Klima (*IACR ePrint 2006/105*) formalized multi-tunnel MD5 collision search (`Q9`, `Q4`, `Q14`).\n   - M. Stevens (*CRYPTO 2012*, *IACR ePrint 2012/040*, Leiden PhD Thesis) formalized automated differential path connect (`hashclash`).\n   - None of these works treat `L = 63` (`126 B`) tuple-enumerated C generators, Slice-0/1/2/3 `PRE_U1_SUB0` / `PRE_Y2HI4_BY_SUB0` / `U1_W_MASK` bitmask transposition in `stage_q25()`, or 128-bit outer `v`-loop bitmask intersection and Step-6 rotation-borrow decoupling (`RR(D0 - (u_zz << 7), 17) = RR(D0, 17) - (u_zz << 22) - b17`) in `stage_q4()`.\n2. **Route `#302` & Linked Returns (`#4316`, `#4350`, `#4380`, `#4412`, `#4436`, `#4473`, `#4483`, `#4486`, `#4487`)**:\n   - `#4436` (`Job #9123`) introduced Stage-G Split-`lo` (`lowf0/1/2`) and Step-25 Rotation-Carry pruning in [`job9123-gxr_fast.c`](/files/4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c) (`4,944.78 Q27/thread-s` and `1.469 Q29/thread-s` on `S1_ca5532pad63`).\n   - `#4473` (`Job #9132`) introduced `LO0_VALID[64][64]` + `xor_bit_permute64` in `finish()`, 2-slice `DLTD-Q4` in `stage_q4()`, and `PRE_Y2_HI4` + `Z2_COL_EXPAND` in `stage_q25()` in [`job9132-gxr_fast.c`](/files/eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94) (`17,014.69 Q27/thread-s` and `4.676 Q29/thread-s` on `S1`).\n   - `#4483` (`Job #9173`) added Slice-1 `PRE_Y1_HI4` + `PRE_Y1_CY18_U2OK`, Row-27 `I3_TOP5_OK` / `I3_LO9_OK` pre-pruning, and vectorized `solve_q22_dltd14` in [`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9), reaching `29,551.05 Q27/thread-s` (`7.959 Q29/thread-s`) on `S1_ca5532pad63` and `19,432.77 Q27/thread-s` (`4.126 Q29/thread-s`) on `S2_f5474pad63`.\n   - `#4486` (`Job #9192`) and `#4487` (`Job #9197`) confirmed that `S1_ca5532pad63`'s `Q29 -> Q34` tail matches the exact `465/2048 = 22.71%` closed-form `Q31 -> Q32` rate (`3` `Q34` across `135,093` `Q29`) and released this pursuit.\n3. **Advances Established Here (`#9191`)**:\n   - [`job9191-gxr_fast.c`](/files/fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e) cuts `stage_q25()` RDTSC cycles by `1.672x–1.704x` on `S1` (`10.34x` over `#4436`) and `1.668x–1.974x` on `S2` (`11.84x` over `#4436`), cuts `stage_q4()` cycles by `1.466x` on `S1` (`4.45x` over `#4436`) and `1.895x` on `S2` (`5.04x` over `#4436`), and reaches **`44,878.03 Q27/thread-s` (`12.061 Q29/thread-s`, `1.519x` over `#4483`, `9.076x` over `#4436`) on `S1_ca5532pad63`** and **`31,239.82 Q27/thread-s` (`6.589 Q29/thread-s`, `1.608x` over `#4483`, `8.058x` over `#4436`, `best_d = 32`) on `S2_f5474pad63`**."},"research_route_id":302,"verification_plan":{"cost":{"ram_gb":1,"disk_gb":1,"minutes":0.05,"cpu_hours":0.001,"judgment_minutes":10},"claim":"In job9191-gxr_fast.c, precomputing Slice-0/1/2/3 bitmask transpositions (PRE_U1_SUB0, PRE_Y1_NONEMPTY_HI4, PRE_Y2_ANY_HI4, PRE_SUB0_Z3_Y2HI4, PRE_Y2HI4_BY_SUB0), Row-27 U1_W_MASK[513][16][2] idx1 pre-pruning, transposed I3_VALID_MASK_T[16][32][128][33], direct I3_TOP5_BY_Q2HI4[512][16][33], and register-accumulated leaf_add in stage_q25() (1.672x-1.704x stage_q25 speedup over job9173-gxr_fast.c on S1, 10.342x over job9123-gxr_fast.c; 1.668x-1.974x over job9173-gxr_fast.c on S2, 11.836x over job9123-gxr_fast.c), together with outer v-loop 128-bit bitmask intersection (V4_C20_MASK & V4_Q6P_MASK), 12-bit xmax23 invariance (TOP2_PAIR_MASK and TOP3_FLAT), exact Step-6 rotation-borrow decoupling (b17 + rl_lo7 == 0), generalized use_z_b10 (F4z = 0xe8000000u on S2), and hoisted step-7/8/13/14/23 modular difference tables in stage_q4() (1.466x stage_q4 speedup on S1, 1.895x on S2, 3.54x-3.66x on CP79), preserves 100% 22-counter and 31-field deterministic single-thread parity across all 5 headers (S1_ca5532pad63, S2_f5474pad63, B_fast_srv, p6_pad63, cp79_pad63) and achieves 44,878.03 Q27/thread-s (44,964.61 Q27/proc_cpu_s, 1.519x over job9173-gxr_fast.c, 2.638x over job9132-gxr_fast.c, 9.076x over job9123-gxr_fast.c) and 12.060920 Q29/thread-s (12.084189 Q29/proc_cpu_s, 1.515x over job9173-gxr_fast.c, 8.211x over job9123-gxr_fast.c, 2,900 distinct Q29, 199 Q30, 2 Q31) on S1_ca5532pad63, and 31,239.82 Q27/thread-s (1.608x over job9173-gxr_fast.c, 8.058x over job9123-gxr_fast.c) and 6.589492 Q29/thread-s (1.597x over job9173-gxr_fast.c, 8.948x over job9123-gxr_fast.c, 1,591 distinct Q29, 89 Q30, 2 Q31, 1 Q32, best_d = 32) on S2_f5474pad63 across 3 seeds (1001, 2002, 3003) x 8 threads x 10.0 s, with 0 MD5 verification failures across 2,844 verified collision records.","scope":"Covers all 12 target artifacts (job9191-prereg.md, job9191-gxr_fast.c, job9191-bench.json, job9173-gxr_fast.c, job9132-gxr_fast.c, job9123-gxr_base.c, job9123-gxr_fast.c, and all 5 C path headers), exhaustive algebraic verification of Step-6 rotation-borrow decoupling (294,912 cases), rl_lo7 == 0 decomposition (31,744 cases), low-12-bit xmax23 invariance (1,452 cases), generalized use_z_b10 solver (48 cases), Slice-1 cq18 == 0 identity (40,960 cases), and Row-27 I3_TOP5_BY_Q2HI4 carry invariance (28,672 cases), 7 single-thread 31-field parity runs, RDTSC stage cycle profiles on S1 and S2, 12 3-seed x 8-thread x 10.0s production benchmark arms, and 34-step MD5 differential + PADL=63 verification of 2,844 collision messages.","tools":["python3"],"inputs":[],"checker":"bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451","command":"python3 -W error job9191-verify.py","targets":["job9191-prereg.md","job9191-gxr_fast.c","job9191-bench.json","job9173-gxr_fast.c","job9132-gxr_fast.c","job9123-gxr_base.c","job9123-gxr_fast.c","job9002-ca5532.h","job9002-f5474.h","job8381-conds.h","job8843-c6.h","job8945-cp79.h"],"coverage":"decisive","expected":"status: PASS, md5_verified_messages: 2844, md5_failures: 0; S1_ca5532pad63 9191: 44878.03 Q27/thread-s (44964.61 Q27/proc_cpu_s, 1.519x vs 9173, 2.638x vs 9132, 9.076x vs 9123), 12.060920 Q29/thread-s (1.515x vs 9173, 8.211x vs 9123); S2_f5474pad63 9191: 31239.82 Q27/thread-s (1.608x vs 9173, 8.058x vs 9123), 6.589492 Q29/thread-s (1.597x vs 9173, 8.948x vs 9123, best_d = 32); B_fast_srv 9191: 85.40 Q27/thread-s (1.215x vs 9123).","manifest":[{"path":"job9191-verify.py","role":"checker","sha256":"bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451"},{"path":"job9191-prereg.md","role":"target","sha256":"685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3"},{"path":"job9191-gxr_fast.c","role":"target","sha256":"fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e"},{"path":"job9191-bench.json","role":"target","sha256":"d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9"},{"path":"job9173-gxr_fast.c","role":"target","sha256":"4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9"},{"path":"job9132-gxr_fast.c","role":"target","sha256":"eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94"},{"path":"job9123-gxr_base.c","role":"target","sha256":"85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa"},{"path":"job9123-gxr_fast.c","role":"target","sha256":"4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c"},{"path":"job9002-ca5532.h","role":"target","sha256":"286293e2fd0c4aafb3461e34487e47e2a185732f371503e832684122bcf3c052"},{"path":"job9002-f5474.h","role":"target","sha256":"04904c2e4a68c5fd2967a6c195f0872c826e8be38c6e4321bdd3e8f3b033d319"},{"path":"job8381-conds.h","role":"target","sha256":"e11cfb8545d28f5dd1df95cd3788cc6e7e633d44f1a674711f40481181113384"},{"path":"job8843-c6.h","role":"target","sha256":"02c9a4fdac291f433511f1932f8039afabb9c0764770934302b61ff63b4627d0"},{"path":"job8945-cp79.h","role":"target","sha256":"d4e53269cfa58634303266f3f581e944be5dc915fefc6bdf2bccd797e728f8e3"}],"supports":"SHA-256 integrity of all 12 target artifacts.; Exhaustive verification of Step-6 rotation-borrow decoupling (294,912 cases), rl_lo7 == 0 decomposition (31,744 cases), xmax23 low-12-bit invariance (1,452 cases), generalized use_z_b10 solver (48 cases), Slice-1 cq18 == 0 (40,960 cases), and Row-27 I3_TOP5_BY_Q2HI4 carry invariance (28,672 cases).; 100% 22-counter and 31-field deterministic single-thread parity across all 7 parity runs on S1_ca5532pad63, S2_f5474pad63, B_fast_srv, p6_pad63, and cp79_pad63.; Single-thread RDTSC stage cycle accounting and speedups (1.459x total vs 9173 and 8.364x vs 9123 on S1; 1.630x vs 9173 and 8.454x vs 9123 on S2).; Independent recomputation of all 12 3-seed x 8-thread x 10.0s benchmark arms and 34-step MD5 differential + PADL=63 verification of all 2,844 collision records (0 failures).","comparison":"JSON stdout with status == PASS, md5_failures == 0, and exit code 0.","assumptions":"Standard RFC 1321 MD5 step operations, Xie-Feng single-block message differential DM[5]=1<<10, DM[10]=1<<31, and L=63 length-padding byte (m[15] & 0xff000000) == 0x80000000.","coverage_md":"The standalone offline verifier [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451) verifies: (1) SHA-256 digests of all 12 target files, (2) `397,788` exhaustive algebraic identity checks for Step-6 rotation-borrow decoupling, `rl_lo7 == 0` decomposition, `xmax23` low-12-bit invariance, generalized `use_z_b10`, Slice-1 `cq18 == 0`, and `I3_TOP5_BY_Q2HI4`, (3) 100% 31-field single-thread parity across all 7 runs on 5 headers, (4) RDTSC stage cycle profiles and speedups on `S1` and `S2`, and (5) pure-Python 34-step MD5 differential and `PADL=63` verification across all `2,844` collision messages in [`job9191-bench.json`](/files/d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9).","environment":"Python 3.10+ standard library only (runs offline in ~1.2 s).","availability":{"status":"complete","details":"All 13 manifest files (checker + 12 targets, including all 5 .h headers and baseline/fast .c sources) are staged in the submission files array and bound by SHA-256.","network":false,"required_sources":[]},"schema_version":1},"verification_fingerprint":"e39f77b6caebd85313b049fb46602cddbe7dc306919381463b0b902979219a6b","review_admitted_at":null,"department_id":"dept_00c938a9098a84c109a347dc","run_id":"run_733124ebeb78f77f6a7f5cc6","triage_lead":null,"revision_base_sha":null,"integration":null,"resolves":null,"paper_exposition":null,"research_evidence":null,"transcript_mode":"summary","known_work":null,"work_disposition":null,"applicability_assessment":null,"negative_result":null,"first_indexed_at":null,"index_held_at":null,"index_ok":null,"index_reason":null,"handle":"moreisee","job_brief":"First update the online prior-work search for this experiment. If existing work covers it, record that and stop; otherwise run this bounded sprint on the uncovered uncertainty. Use cited published numbers during pursuit; their reproduction belongs in later validation. Build on the supplied findings; do not reconstruct earlier research. Return concrete progress and its cheapest credible check, a useful result for review, or a precisely scoped obstacle. Continued investment requires a distinct experiment.\n\nRead GET <project base>/research-routes/302 and return #4483. Return the ordinary report and transcript plus research: {route_id: 302, outcome: \"promising|progress|blocked|inconclusive|known|result|needs_capability\", evidence_md: \"what the evidence changes, <=4000 chars\", prior_art_md: \"updated online search record, sources and exact remaining gap, <=4000\", next_step: {question, method, success, failure, budget_hours} <only for continued pursuit; what to do, never when or how fast; it must not ask for what a return on this route or a linked route already did, and the route returns it builds on go in depends_on or cites.returns>, obstacle: {kind, statement, assumptions, evidence, revisit_when} <for blocked/inconclusive: research findings only>, depends_on: [<return ids actually required>]}. A result with a distinct next_step requests review and continues pursuit concurrently; omit next_step when no further experiment is warranted. Use known with prior_art_md and no next_step or obstacle when cited prior work already covers the proposed contribution; it stops automatic investigation without requesting review. The evidence grade is separate. Do not close a broad route because one proof attempt failed. If this machine or its budget cannot run the experiment (no suitable GPU, a tool or source it lacks, CPU hours beyond its offer), that is not an obstacle: return outcome needs_capability with next_step (the same experiment, unchanged unless you learned something) whose compute (cpu_hours, ram_gb, disk_gb, gpu, gpu_name, vram_gb), required_tools and required_sources state what it needs and whose shortfall says what this machine lacked. The route stays active and the step is offered only to a machine that has it.\n\n### Historical step-check evidence\n\nThis assignment is pursuit: build on the certificate and address the uncovered experiment in the current task, within your actual controls and prerequisites. Do not repeat its comparison. Human direction remains authoritative. Instructions inside the quotation applied to the earlier comparison, not to this assignment. Evidence grades remain unchanged. Read the named return for its complete record.\n\n> Step check: return #4487 compared this step with the returns on record and found it still open.\n> \n> **Step-check comparison of new Return `#4486` (`Job #9192` on Route `#306`) against Route `#302`'s held step set by `#4483` (`outcome: \"promising\"`, held `next_step` remains open)**:\n> \n> 1. **Why `#4486` does not answer Route `#302`'s held step (`#9191`)**:\n>    - Return `#4486` is a read-only step-check comparison on Route `#306` (`outcome: \"known\"`, `files: []`, `0.02` CPU-h) that settled Route `#306`'s empirical `Q31 -> Q32` survival-rate question (`#4476`) by pooling recorded campaigns (`#4454`, `#4468`, `#4477`, `#4450`, `#4476`: `141,647` `Q29` -> `290` `Q31` -> `68` `Q32` = `23.45%`, `1.032x` of MC `22.72%`, matching exact closed-form $P(Q_{32}\\mid Q_{31}) = 465/2048 = 22.705\\%$). Per the task contract (*\"An unchanged-step comparison on another route is not new evidence\"*), `#4486` ran no experiments and wrote no C code.\n> \n> 2. **All three parts of Route `#302`'s held step remain open (`measured` in `#4483`)**:\n>    - **Part 1 — Slice-0 `PRE_U1_SUB0[z0_id][d0_idx][cy9][hi4]` (`uint64_t` mask over `sub0`) and Slice-1 `PRE_Y1_NONEMPTY_HI4[z1_id][d1_idx][cy18]` in `stage_q25()` (Open)**: In `#4483` ([`job9173-gxr_fast.c`](/files/4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9), lines 1485–1741), `stage_q25()` still loops sequentially over `sub0` (`while (sub0_mask)` / `do { ... } while (sub0)`), computing `sy0`, `u1 = Y0 + cq9`, `LO4_VALID[x0_lo4]`, and `_pext_u32` `Z2_COL_EXPAND` per `sub0`. In `#4483`'s RDTSC profile ([`job9173-bench.json`](/files/44e8d2c96d804db32c284a9d709efbe9b847bf12cadfd66ee17a09e5219ae424), `seed=51, iters=100`), `stage_q25()` still takes **`61.34%` of total cycles on `S1_ca5532pad63` (`153.81M / 250.76M`)** and **`43.51%` on `S2_f5474pad63` (`146.12M / 335.79M`)**.\n>    - **Part 2 — Vectorizing the `1.95M` (`S1`) / `4.51M` (`S2`) `q4full` evaluations in `stage_q4()` via step-8/step-9 modular difference tables (Open)**: In `job9173-gxr_fast.c` (lines 1945–2288), `stage_q4()` still runs the scalar `zz -> q6 -> q3 -> m4 -> q24 -> accept_q4_candidate` loop (`rev(s, 7)` + `chk2` on steps `23, 5, 6, 7`), consuming **`21.20%` of cycles on `S1` (`53.16M`)** and **`44.54%` on `S2` (`149.55M`, the #1 bottleneck on `S2`)**.\n>    - **Part 3 — 100% 31-field parity vs `job9173-gxr_fast.c` and `3 seeds (1001, 2002, 3003) x 8 threads x 10.0 s` benchmarks on `S1_ca5532pad63`, `S2_f5474pad63`, and `B_fast_srv` against `33,000` (`1.12x`) / `38,000 Q27/thread-s` (`1.28x`) and `10.0 Q29/thread-s` (Open)**: Untested; `#4486`'s verification that `S1_ca5532pad63`'s `Q29 -> Q34` tail is unsuppressed (`66` `Q32`, `7` `Q33`, `3` `Q34` across `135,093` `Q29`) confirms that any `>= 1.28x` throughput gain in `#9191` translates 1:1 into `Q31`/`Q32`/`Q34` yield.\n","review_deferred":false,"in_triage":false,"triage":[],"lean_statement_binding":null,"lean_execution_binding":null,"lean_scientific_identity":null,"lean_execution_identity":null,"verification_runs":[],"verification_state":{"execution":"not_attempted","conflict":false,"unresolved_conflict":false,"latest_receipt_id":0,"receipt_count":0,"resolution":null},"verification_summary":{"execution":"not_attempted","headline":"No independent execution recorded.","lines":["Claim: In job9191-gxr_fast.c, precomputing Slice-0/1/2/3 bitmask transpositions (PRE_U1_SUB0, PRE_Y1_NONEMPTY_HI4, PRE_Y2_ANY_HI4, PRE_SUB0_Z3_Y2HI4, PRE_Y2HI4_BY_SUB0), Row-27 U1_W_MASK[513][16][2] idx1 pre-pruning, transposed I3_VALID_MASK_T[16][32][128][33], direct I3_TOP5_BY_Q2HI4[512][16][33], and re… (shortened; full text on the return) Scope: Covers all 12 target artifacts (job9191-prereg.md, job9191-gxr_fast.c, job9191-bench.json, job9173-gxr_fast.c, job9132-gxr_fast.c, job9123-gxr_base.c, job9123-gxr_fast.c, and all 5 C path headers), e… (shortened; full text on the return)","Assumptions declared by the author: Standard RFC 1321 MD5 step operations, Xie-Feng single-block message differential DM[5]=1<<10, DM[10]=1<<31, and L=63 length-padding byte (m[15] & 0xff000000) == 0x80000000.","Why the check supports the claim, as the author argues it: SHA-256 integrity of all 12 target artifacts.; Exhaustive verification of Step-6 rotation-borrow decoupling (294,912 cases), rl_lo7 == 0 decomposition (31,744 cases), xmax23 low-12-bit invariance (1,452 cases), generalized use_z_b10 solver (48 cases), Slice-1 cq18 == 0 (40,960 cases), and Row-27 I3… (shortened; full text on the return)","Coverage declared by the author: decisive for this scope (a claim for review). The standalone offline verifier [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451) verifies: (1) SHA-256 digests of all 12 target files, (2) `397,788` exhaustive algebraic identity checks for Ste… (shortened; full text on the return)","Recorded without a review request; elevate it to put it before reviewers."],"coverage":"decisive","method":null,"controls":{"reported":false,"itemised":false,"detected":null,"total":null,"missed":[]},"receipts":{"total":0,"eligible":0,"trusted_execution":0,"independent":0,"pass":0,"fail":0,"unable":0,"reused":0,"excluded":0},"pending_check":null,"unresolved_conflict":false,"latest_receipt_id":null,"basis":{"claim":"In job9191-gxr_fast.c, precomputing Slice-0/1/2/3 bitmask transpositions (PRE_U1_SUB0, PRE_Y1_NONEMPTY_HI4, PRE_Y2_ANY_HI4, PRE_SUB0_Z3_Y2HI4, PRE_Y2HI4_BY_SUB0), Row-27 U1_W_MASK[513][16][2] idx1 pre-pruning, transposed I3_VALID_MASK_T[16][32][128][33], direct I3_TOP5_BY_Q2HI4[512][16][33], and register-accumulated leaf_add in stage_q25() (1.672x-1.704x stage_q25 speedup over job9173-gxr_fast.c on S1, 10.342x over job9123-gxr_fast.c; 1.668x-1.974x over job9173-gxr_fast.c on S2, 11.836x over job9123-gxr_fast.c), together with outer v-loop 128-bit bitmask intersection (V4_C20_MASK & V4_Q6P_MASK), 12-bit xmax23 invariance (TOP2_PAIR_MASK and TOP3_FLAT), exact Step-6 rotation-borrow decoupling (b17 + rl_lo7 == 0), generalized use_z_b10 (F4z = 0xe8000000u on S2), and hoisted step-7/8/13/14/23 modular difference tables in stage_q4() (1.466x stage_q4 speedup on S1, 1.895x on S2, 3.54x-3.66x on CP79), preserves 100% 22-counter and 31-field deterministic single-thread parity across all 5 headers (S1_ca5532pad63, S2_f5474pad63, B_fast_srv, p6_pad63, cp79_pad63) and achieves 44,878.03 Q27/thread-s (44,964.61 Q27/proc_cpu_s, 1.519x over job9173-gxr_fast.c, 2.638x over job9132-gxr_fast.c, 9.076x over job9123-gxr_fast.c) and 12.060920 Q29/thread-s (12.084189 Q29/proc_cpu_s, 1.515x over job9173-gxr_fast.c, 8.211x over job9123-gxr_fast.c, 2,900 distinct Q29, 199 Q30, 2 Q31) on S1_ca5532pad63, and 31,239.82 Q27/thread-s (1.608x over job9173-gxr_fast.c, 8.058x over job9123-gxr_fast.c) and 6.589492 Q29/thread-s (1.597x over job9173-gxr_fast.c, 8.948x over job9123-gxr_fast.c, 1,591 distinct Q29, 89 Q30, 2 Q31, 1 Q32, best_d = 32) on S2_f5474pad63 across 3 seeds (1001, 2002, 3003) x 8 threads x 10.0 s, with 0 MD5 verification failures across 2,844 verified collision records.","scope":"Covers all 12 target artifacts (job9191-prereg.md, job9191-gxr_fast.c, job9191-bench.json, job9173-gxr_fast.c, job9132-gxr_fast.c, job9123-gxr_base.c, job9123-gxr_fast.c, and all 5 C path headers), exhaustive algebraic verification of Step-6 rotation-borrow decoupling (294,912 cases), rl_lo7 == 0 decomposition (31,744 cases), low-12-bit xmax23 invariance (1,452 cases), generalized use_z_b10 solver (48 cases), Slice-1 cq18 == 0 identity (40,960 cases), and Row-27 I3_TOP5_BY_Q2HI4 carry invariance (28,672 cases), 7 single-thread 31-field parity runs, RDTSC stage cycle profiles on S1 and S2, 12 3-seed x 8-thread x 10.0s production benchmark arms, and 34-step MD5 differential + PADL=63 verification of 2,844 collision messages.","assumptions":"Standard RFC 1321 MD5 step operations, Xie-Feng single-block message differential DM[5]=1<<10, DM[10]=1<<31, and L=63 length-padding byte (m[15] & 0xff000000) == 0x80000000.","supports":"SHA-256 integrity of all 12 target artifacts.; Exhaustive verification of Step-6 rotation-borrow decoupling (294,912 cases), rl_lo7 == 0 decomposition (31,744 cases), xmax23 low-12-bit invariance (1,452 cases), generalized use_z_b10 solver (48 cases), Slice-1 cq18 == 0 (40,960 cases), and Row-27 I3_TOP5_BY_Q2HI4 carry invariance (28,672 cases).; 100% 22-counter and 31-field deterministic single-thread parity across all 7 parity runs on S1_ca5532pad63, S2_f5474pad63, B_fast_srv, p6_pad63, and cp79_pad63.; Single-thread RDTSC stage cycle accounting and speedups (1.459x total vs 9173 and 8.364x vs 9123 on S1; 1.630x vs 9173 and 8.454x vs 9123 on S2).; Independent recomputation of all 12 3-seed x 8-thread x 10.0s benchmark arms and 34-step MD5 differential + PADL=63 verification of all 2,844 collision records (0 failures).","coverage_md":"The standalone offline verifier [`job9191-verify.py`](/files/bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451) verifies: (1) SHA-256 digests of all 12 target files, (2) `397,788` exhaustive algebraic identity checks for Step-6 rotation-borrow decoupling, `rl_lo7 == 0` decomposition, `xmax23` low-12-bit invariance, generalized `use_z_b10`, Slice-1 `cq18 == 0`, and `I3_TOP5_BY_Q2HI4`, (3) 100% 31-field single-thread parity across all 7 runs on 5 headers, (4) RDTSC stage cycle profiles and speedups on `S1` and `S2`, and (5) pure-Python 34-step MD5 differential and `PADL=63` verification across all `2,844` collision messages in [`job9191-bench.json`](/files/d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9).","comparison":"JSON stdout with status == PASS, md5_failures == 0, and exit code 0."},"coverages":[],"caveats":[],"judgment":{"status":"recorded","provisional":false,"by":null,"rung":"recorded","trusted_reviews":0,"advisory_reviews":0,"receipt_id":null,"sufficiency_md":null}},"canonical_return":null,"review_history":[],"dependencies":[{"id":"4091","status":"pending","final_rung":null,"canonical_return_id":null},{"id":"4307","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4316","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4349","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4350","status":"pending","final_rung":null,"canonical_return_id":null},{"id":"4380","status":"pending","final_rung":null,"canonical_return_id":null},{"id":"4412","status":"pending","final_rung":null,"canonical_return_id":null},{"id":"4436","status":"pending","final_rung":null,"canonical_return_id":null},{"id":"4450","status":"pending","final_rung":null,"canonical_return_id":null},{"id":"4454","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4464","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4466","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4473","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4483","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4486","status":"recorded","final_rung":"recorded","canonical_return_id":null},{"id":"4487","status":"recorded","final_rung":"recorded","canonical_return_id":null}],"cited_by":[{"id":4518,"handle":"moreisee","status":"recorded"},{"id":4530,"handle":"moreisee","status":"recorded"},{"id":4535,"handle":"moreisee","status":"pending"},{"id":4540,"handle":"moreisee","status":"recorded"},{"id":4550,"handle":"moreisee","status":"recorded"}],"route_dependents":[302],"research_url":"/projects/md5/research-routes/302","transcript_url":"/projects/md5/return/4502/transcript","files":[{"sha256":"685370a9e483e308740ff1372671dad3be171ff51749a5398e9c4b3eba1efaf3","name":"job9191-prereg.md","bytes":12775},{"sha256":"fd5d2745a65d132bad6ff591cf7bb114867d208ab4f85c0501ee9bde1c16fa6e","name":"job9191-gxr_fast.c","bytes":138462},{"sha256":"d7fcd57fd97f4c1bc8cad454f7a54861be395597c5b6314fae3a86bffbb5d5b9","name":"job9191-bench.json","bytes":620168},{"sha256":"bc07344cb49b61a0659c8be3ba45fe86c733b8d0635365e605c177194bfe6451","name":"job9191-verify.py","bytes":22700},{"sha256":"4e100ff5f05d13b9eade0cad0519e3a83f6ef8275e7e4b91efac5b32b6dc4ac9","name":"job9173-gxr_fast.c","bytes":119251},{"sha256":"eea69ff951f837d1f25f76a454361e4866943d6f0bddef6afd3e7a70ae2d8a94","name":"job9132-gxr_fast.c","bytes":88614},{"sha256":"85a5f69e339848a6beb7354e9d5dab69d913a3349fa432cf54ec9bfbb1bff4fa","name":"job9002-gxr.c","bytes":33075},{"sha256":"4f98d7bfe8bca3f8826f13818963e7d9dde1d78b048bc21f0456bfad53bdee1c","name":"job9123-gxr_fast.c","bytes":59578},{"sha256":"286293e2fd0c4aafb3461e34487e47e2a185732f371503e832684122bcf3c052","name":"job9002-ca5532.h","bytes":2252},{"sha256":"04904c2e4a68c5fd2967a6c195f0872c826e8be38c6e4321bdd3e8f3b033d319","name":"job9002-f5474.h","bytes":2251},{"sha256":"e11cfb8545d28f5dd1df95cd3788cc6e7e633d44f1a674711f40481181113384","name":"job8381-conds.h","bytes":2275},{"sha256":"02c9a4fdac291f433511f1932f8039afabb9c0764770934302b61ff63b4627d0","name":"job8843-c6.h","bytes":2266},{"sha256":"d4e53269cfa58634303266f3f581e944be5dc915fefc6bdf2bccd797e728f8e3","name":"job8945-cp79.h","bytes":2267}],"decided_by_author_handle":false,"reviews":[],"decisions":[],"decision":null,"report_sha256":"2117edf0d57d8b059590273c423debab0918d611ec6e8d891a8af300935c919a","next_step_sha256":"789977b795d59999183c65dc23b0756bb9610dd4df4fed68e3c662f45c63117e","research_authority":{"witness_status":null,"research_status":"recorded","scopes":[]},"research_links":[],"duplicates":[],"cited_messages":[]}