Problem 369
Badugi: count n-card hands (4≤n≤13) containing a 4-suit matching.
View problem on Project Euler
Performance comparison
| Metric | Our solution | Best known |
|---|
| Time complexity | O(n^3) | O(n) |
|---|
| Space complexity | O(n^2) | O(1) |
|---|
| Approach | Flow solution | Direct hand evaluation or enumeration |
|---|
| Verdict | Suboptimal |
|---|
Flow source
# Project Euler 369
# Badugi: count n-card hands (4≤n≤13) containing a 4-suit matching.
extern {
function calloc(n: i64, size: i64) -> ptr<void>
function free(p: ptr<void>) -> void
}
function popcount4(x: i32) -> i32 {
return (x & 1) + ((x >> 1) & 1) + ((x >> 2) & 1) + ((x >> 3) & 1)
}
function update_reachable(reach: i32, suits_present: i32) -> i32 {
let mut new_reach: i32 = reach
let mut m: i32 = 0
while m < 16 {
if ((reach >> m) & 1) == 1 {
let mut b: i32 = suits_present & (0xF ^ m)
while b != 0 {
let lsb: i32 = b & (0 - b)
new_reach = new_reach | (1 << (m | lsb))
b = b - lsb
}
}
m = m + 1
}
return new_reach & 65535
}
function main() -> i32 {
let max_n: i32 = 13
# BFS reachable reach-bitsets
let states: ptr<i32> = calloc(128, 4)
let index_of: ptr<i32> = calloc(65536, 4)
if states == null || index_of == null { return 1 }
let mut i: i32 = 0
while i < 65536 {
index_of[i] = -1
i = i + 1
}
let start: i32 = 1
states[0] = start
index_of[start] = 0
let mut nstates: i32 = 1
let mut qi: i32 = 0
while qi < nstates {
let reach: i32 = states[qi]
let mut suits: i32 = 0
while suits < 16 {
let nxt: i32 = update_reachable(reach, suits)
if index_of[nxt] < 0 {
index_of[nxt] = nstates
states[nstates] = nxt
nstates = nstates + 1
}
suits = suits + 1
}
qi = qi + 1
}
let trans: ptr<i32> = calloc((nstates * 16) as i64, 4)
if trans == null { return 1 }
i = 0
while i < nstates {
let mut suits2: i32 = 0
while suits2 < 16 {
trans[i * 16 + suits2] = index_of[update_reachable(states[i], suits2)]
suits2 = suits2 + 1
}
i = i + 1
}
let dp: ptr<i64> = calloc((nstates * (max_n + 1)) as i64, 8)
let ndp: ptr<i64> = calloc((nstates * (max_n + 1)) as i64, 8)
if dp == null || ndp == null { return 1 }
dp[0 * (max_n + 1) + 0] = 1 # start_idx is 0
let mut rank: i32 = 0
while rank < 13 {
let mut z: i64 = 0
while z < (nstates * (max_n + 1)) as i64 {
ndp[z] = 0
z = z + 1
}
i = 0
while i < nstates {
let mut k: i32 = 0
let base: i32 = i * (max_n + 1)
let mut any: i32 = 0
while k <= max_n {
if dp[base + k] != 0 { any = 1; break }
k = k + 1
}
if any == 1 {
let mut suits3: i32 = 0
while suits3 < 16 {
let j: i32 = trans[i * 16 + suits3]
let add: i32 = popcount4(suits3)
k = 0
while k <= max_n {
let v: i64 = dp[base + k]
let nk: i32 = k + add
if v != 0 && nk <= max_n {
ndp[j * (max_n + 1) + nk] = ndp[j * (max_n + 1) + nk] + v
}
k = k + 1
}
suits3 = suits3 + 1
}
}
i = i + 1
}
z = 0
while z < (nstates * (max_n + 1)) as i64 {
dp[z] = ndp[z]
z = z + 1
}
rank = rank + 1
}
let full_bit: i32 = 1 << 15
let mut ans: i64 = 0
i = 0
while i < nstates {
if (states[i] & full_bit) != 0 {
let mut n: i32 = 4
while n <= max_n {
ans = ans + dp[i * (max_n + 1) + n]
n = n + 1
}
}
i = i + 1
}
printf("%lld\n", ans)
free(ndp)
free(dp)
free(trans)
free(index_of)
free(states)
return 0
}
Generated C
#include <stdint.h>
#include <stdbool.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
/* Flow runtime helpers */
typedef struct flow_temp_node { struct flow_temp_node* next; } flow_temp_node;
static flow_temp_node* flow_temp_head = NULL;
static int flow_temp_atexit_set = 0;
__attribute__((unused)) static void flow_temp_free_all(void) {
while (flow_temp_head) {
flow_temp_node* n = flow_temp_head;
flow_temp_head = n->next;
free(n);
}
}
__attribute__((unused)) static void* flow_temp_alloc(size_t nbytes) {
flow_temp_node* node = (flow_temp_node*)malloc(sizeof(flow_temp_node) + nbytes);
if (!node) return NULL;
node->next = flow_temp_head;
flow_temp_head = node;
if (!flow_temp_atexit_set) {
flow_temp_atexit_set = 1;
atexit(flow_temp_free_all);
}
return (void*)(node + 1);
}
#ifndef FLOW_DIAG
#define FLOW_DIAG(msg) fprintf(stderr, "%s", (msg))
#endif
#ifndef FLOW_LOG
#define FLOW_LOG(fmt, ...) printf(fmt, __VA_ARGS__)
#endif
#ifndef FLOW_LOG_EMPTY
#define FLOW_LOG_EMPTY(fmt) printf(fmt)
#endif
static char* flow_strcat(const char* a, const char* b) {
size_t la = strlen(a ? a : ""), lb = strlen(b ? b : "");
char* r = (char*)flow_temp_alloc(la + lb + 1);
if (!r) return NULL;
if (la) memcpy(r, a, la);
if (lb) memcpy(r + la, b, lb);
r[la + lb] = '\0';
return r;
}
#define __flow_in_arr(arr, val) __extension__ ({ \
int _found = 0; \
size_t _n = sizeof(arr)/sizeof((arr)[0]); \
for (size_t _i = 0; _i < _n; _i++) { \
if ((arr)[_i] == (val)) { _found = 1; break; } \
} _found; })
/* Unified fault handler (MISRA #279) — override with -DFLOW_FAULT_HANDLER=fn */
#ifndef FLOW_FAULT_HANDLER
__attribute__((unused)) static inline void flow_fault_handler(const char* msg) {
fprintf(stderr, "flow: %s\n", msg ? msg : "fault");
abort();
#if defined(__GNUC__) || defined(__clang__)
__builtin_unreachable();
#endif
}
#else
#define flow_fault_handler FLOW_FAULT_HANDLER
#endif
#define flow_div_by_zero_handler() flow_fault_handler("division by zero")
#define flow_shift_ub_handler() flow_fault_handler("invalid shift (amount out of range or left-shift of negative)")
#ifndef FLOW_CHECKED_DIV
#define FLOW_CHECKED_DIV(L, R) (((R) != 0) ? ((L) / (R)) : (flow_div_by_zero_handler(), (L) * 0))
#endif
#ifndef FLOW_CHECKED_MOD
#define FLOW_CHECKED_MOD(L, R) (((R) != 0) ? ((L) % (R)) : (flow_div_by_zero_handler(), (L) * 0))
#endif
#ifndef FLOW_CHECKED_SHL
#define FLOW_CHECKED_SHL(L, R) ((((R) >= 0) && ((unsigned long long)(R) < (sizeof(L) * 8ull)) && ((L) >= 0)) ? ((L) << (R)) : (flow_shift_ub_handler(), (L) * 0))
#endif
#ifndef FLOW_CHECKED_SHR
#define FLOW_CHECKED_SHR(L, R) ((((R) >= 0) && ((unsigned long long)(R) < (sizeof(L) * 8ull))) ? ((L) >> (R)) : (flow_shift_ub_handler(), (L) * 0))
#endif
#include <math.h>
void* _ui_state = NULL;
static inline float i32_to_f32(int32_t v) { return (float)v; }
/* Host stub for @gpu kernels (device codegen replaces this). */
static inline int32_t gpu_thread_id(void) { return 0; }
int32_t popcount4_i32(int32_t x);
int32_t update_reachable_i32_i32(int32_t reach, int32_t suits_present);
int32_t main(void);
int32_t popcount4_i32(int32_t x) {
return ((((x & 1) + (FLOW_CHECKED_SHR((x), (1)) & 1)) + (FLOW_CHECKED_SHR((x), (2)) & 1)) + (FLOW_CHECKED_SHR((x), (3)) & 1));
}
int32_t update_reachable_i32_i32(int32_t reach, int32_t suits_present) {
int32_t new_reach = reach;
int32_t m = 0;
while (m < 16) {
if ((FLOW_CHECKED_SHR((reach), (m)) & 1) == 1) {
int32_t b = (suits_present & (15 ^ m));
while (b != 0) {
int32_t lsb = (b & (0 - b));
new_reach = (new_reach | FLOW_CHECKED_SHL((1), ((m | lsb))));
b = (b - lsb);
}
}
m = (m + 1);
}
return (new_reach & 65535);
}
int32_t main(void) {
int32_t max_n = 13;
int32_t* states = (int32_t*)(calloc(128, 4));
int32_t* index_of = (int32_t*)(calloc(65536, 4));
if ((states == NULL || index_of == NULL)) {
return 1;
}
int32_t i = 0;
while (i < 65536) {
index_of[i] = (-1);
i = (i + 1);
}
int32_t start = 1;
states[0] = start;
index_of[start] = 0;
int32_t nstates = 1;
int32_t qi = 0;
while (qi < nstates) {
int32_t reach = states[qi];
int32_t suits = 0;
while (suits < 16) {
int32_t nxt = update_reachable_i32_i32(reach, suits);
if (index_of[nxt] < 0) {
index_of[nxt] = nstates;
states[nstates] = nxt;
nstates = (nstates + 1);
}
suits = (suits + 1);
}
qi = (qi + 1);
}
int32_t* trans = (int32_t*)(calloc(((int64_t)((nstates * 16))), 4));
if (trans == NULL) {
return 1;
}
i = 0;
while (i < nstates) {
int32_t suits2 = 0;
while (suits2 < 16) {
trans[((i * 16) + suits2)] = index_of[update_reachable_i32_i32(states[i], suits2)];
suits2 = (suits2 + 1);
}
i = (i + 1);
}
int64_t* dp = (int64_t*)(calloc(((int64_t)((nstates * (max_n + 1)))), 8));
int64_t* ndp = (int64_t*)(calloc(((int64_t)((nstates * (max_n + 1)))), 8));
if ((dp == NULL || ndp == NULL)) {
return 1;
}
dp[((0 * (max_n + 1)) + 0)] = 1;
int32_t rank = 0;
while (rank < 13) {
int64_t z = 0;
while (z < ((int64_t)((nstates * (max_n + 1))))) {
ndp[z] = 0;
z = (z + 1);
}
i = 0;
while (i < nstates) {
int32_t k = 0;
int32_t base = (i * (max_n + 1));
int32_t any = 0;
while (k <= max_n) {
if (dp[(base + k)] != 0) {
any = 1;
break;
}
k = (k + 1);
}
if (any == 1) {
int32_t suits3 = 0;
while (suits3 < 16) {
int32_t j = trans[((i * 16) + suits3)];
int32_t add = popcount4_i32(suits3);
k = 0;
while (k <= max_n) {
int64_t v = dp[(base + k)];
int32_t nk = (k + add);
if ((v != 0 && nk <= max_n)) {
ndp[((j * (max_n + 1)) + nk)] = (ndp[((j * (max_n + 1)) + nk)] + v);
}
k = (k + 1);
}
suits3 = (suits3 + 1);
}
}
i = (i + 1);
}
z = 0;
while (z < ((int64_t)((nstates * (max_n + 1))))) {
dp[z] = ndp[z];
z = (z + 1);
}
rank = (rank + 1);
}
int32_t full_bit = FLOW_CHECKED_SHL((1), (15));
int64_t ans = 0;
i = 0;
while (i < nstates) {
if ((states[i] & full_bit) != 0) {
int32_t n = 4;
while (n <= max_n) {
ans = (ans + dp[((i * (max_n + 1)) + n)]);
n = (n + 1);
}
}
i = (i + 1);
}
printf("%lld\n", ans);
free(ndp);
free(dp);
free(trans);
free(index_of);
free(states);
return 0;
}
Generated MLIR
module {
llvm.func @printf(!llvm.ptr, ...) -> i32
llvm.mlir.global internal constant @str_0("%lld\n\00") {addr_space = 0 : i32} : !llvm.array<6 x i8>
func.func private @calloc(i64, i64) -> !llvm.ptr
func.func private @free(!llvm.ptr) -> ()
func.func @popcount4(%arg0: i32) -> i32 {
%0 = arith.constant 1 : i32
%1 = arith.andi %arg0, %0 : i32
%2 = arith.constant 1 : i32
%3 = arith.shrsi %arg0, %2 : i32
%4 = arith.constant 1 : i32
%5 = arith.andi %3, %4 : i32
%6 = arith.addi %1, %5 : i32
%7 = arith.constant 2 : i32
%8 = arith.shrsi %arg0, %7 : i32
%9 = arith.constant 1 : i32
%10 = arith.andi %8, %9 : i32
%11 = arith.addi %6, %10 : i32
%12 = arith.constant 3 : i32
%13 = arith.shrsi %arg0, %12 : i32
%14 = arith.constant 1 : i32
%15 = arith.andi %13, %14 : i32
%16 = arith.addi %11, %15 : i32
func.return %16 : i32
}
func.func @update_reachable(%arg0: i32, %arg1: i32) -> i32 {
%17 = llvm.mlir.constant(1 : i64) : i64
%18 = llvm.alloca %17 x i32 : (i64) -> !llvm.ptr
llvm.store %arg0, %18 : i32, !llvm.ptr
%19 = arith.constant 0 : i32
%20 = llvm.mlir.constant(1 : i64) : i64
%21 = llvm.alloca %20 x i32 : (i64) -> !llvm.ptr
llvm.store %19, %21 : i32, !llvm.ptr
cf.br ^bb0
^bb0:
%22 = llvm.load %21 : !llvm.ptr -> i32
%23 = arith.constant 16 : i32
%24 = arith.cmpi slt, %22, %23 : i32
cf.cond_br %24, ^bb1, ^bb2
^bb1:
%25 = llvm.load %21 : !llvm.ptr -> i32
%26 = arith.shrsi %arg0, %25 : i32
%27 = arith.constant 1 : i32
%28 = arith.andi %26, %27 : i32
%29 = arith.constant 1 : i32
%30 = arith.cmpi eq, %28, %29 : i32
cf.cond_br %30, ^bb3, ^bb4
^bb3:
%31 = arith.constant 15 : i32
%32 = llvm.load %21 : !llvm.ptr -> i32
%33 = arith.xori %31, %32 : i32
%34 = arith.andi %arg1, %33 : i32
%35 = llvm.mlir.constant(1 : i64) : i64
%36 = llvm.alloca %35 x i32 : (i64) -> !llvm.ptr
llvm.store %34, %36 : i32, !llvm.ptr
cf.br ^bb6
^bb6:
%37 = llvm.load %36 : !llvm.ptr -> i32
%38 = arith.constant 0 : i32
%39 = arith.cmpi ne, %37, %38 : i32
cf.cond_br %39, ^bb7, ^bb8
^bb7:
%40 = llvm.load %36 : !llvm.ptr -> i32
%41 = arith.constant 0 : i32
%42 = llvm.load %36 : !llvm.ptr -> i32
%43 = arith.subi %41, %42 : i32
%44 = arith.andi %40, %43 : i32
%45 = llvm.load %18 : !llvm.ptr -> i32
%46 = arith.constant 1 : i32
%47 = llvm.load %21 : !llvm.ptr -> i32
%48 = arith.ori %47, %44 : i32
%49 = arith.shli %46, %48 : i32
%50 = arith.ori %45, %49 : i32
llvm.store %50, %18 : i32, !llvm.ptr
%51 = llvm.load %36 : !llvm.ptr -> i32
%52 = arith.subi %51, %44 : i32
llvm.store %52, %36 : i32, !llvm.ptr
cf.br ^bb6
^bb8:
cf.br ^bb5
^bb4:
cf.br ^bb5
^bb5:
%53 = llvm.load %21 : !llvm.ptr -> i32
%54 = arith.constant 1 : i32
%55 = arith.addi %53, %54 : i32
llvm.store %55, %21 : i32, !llvm.ptr
cf.br ^bb0
^bb2:
%56 = llvm.load %18 : !llvm.ptr -> i32
%57 = arith.constant 65535 : i32
%58 = arith.andi %56, %57 : i32
func.return %58 : i32
}
func.func @main() -> i32 {
%59 = arith.constant 13 : i32
%61 = arith.constant 128 : i32
%62 = arith.constant 4 : i32
%63 = arith.extsi %61 : i32 to i64
%64 = arith.extsi %62 : i32 to i64
%60 = func.call @calloc(%63, %64) : (i64, i64) -> !llvm.ptr
%66 = arith.constant 65536 : i32
%67 = arith.constant 4 : i32
%68 = arith.extsi %66 : i32 to i64
%69 = arith.extsi %67 : i32 to i64
%65 = func.call @calloc(%68, %69) : (i64, i64) -> !llvm.ptr
%70 = llvm.mlir.zero : !llvm.ptr
%71 = llvm.icmp "eq" %60, %70 : !llvm.ptr
%72 = scf.if %71 -> (i1) {
%73 = arith.constant true
scf.yield %73 : i1
} else {
%74 = llvm.mlir.zero : !llvm.ptr
%75 = llvm.icmp "eq" %65, %74 : !llvm.ptr
scf.yield %75 : i1
}
cf.cond_br %72, ^bb9, ^bb10
^bb9:
%76 = arith.constant 1 : i32
func.return %76 : i32
^bb10:
cf.br ^bb11
^bb11:
%77 = arith.constant 0 : i32
%78 = llvm.mlir.constant(1 : i64) : i64
%79 = llvm.alloca %78 x i32 : (i64) -> !llvm.ptr
llvm.store %77, %79 : i32, !llvm.ptr
cf.br ^bb12
^bb12:
%80 = llvm.load %79 : !llvm.ptr -> i32
%81 = arith.constant 65536 : i32
%82 = arith.cmpi slt, %80, %81 : i32
cf.cond_br %82, ^bb13, ^bb14
^bb13:
%83 = arith.constant 1 : i32
%85 = arith.constant 0 : i32
%84 = arith.subi %85, %83 : i32
%86 = llvm.load %79 : !llvm.ptr -> i32
%87 = arith.extsi %86 : i32 to i64
%88 = llvm.getelementptr %65[%87] : (!llvm.ptr, i64) -> !llvm.ptr, i32
llvm.store %84, %88 : i32, !llvm.ptr
%89 = llvm.load %79 : !llvm.ptr -> i32
%90 = arith.constant 1 : i32
%91 = arith.addi %89, %90 : i32
llvm.store %91, %79 : i32, !llvm.ptr
cf.br ^bb12
^bb14:
%92 = arith.constant 1 : i32
%93 = arith.constant 0 : i32
%94 = arith.extsi %93 : i32 to i64
%95 = llvm.getelementptr %60[%94] : (!llvm.ptr, i64) -> !llvm.ptr, i32
llvm.store %92, %95 : i32, !llvm.ptr
%96 = arith.constant 0 : i32
%97 = arith.extsi %92 : i32 to i64
%98 = llvm.getelementptr %65[%97] : (!llvm.ptr, i64) -> !llvm.ptr, i32
llvm.store %96, %98 : i32, !llvm.ptr
%99 = arith.constant 1 : i32
%100 = llvm.mlir.constant(1 : i64) : i64
%101 = llvm.alloca %100 x i32 : (i64) -> !llvm.ptr
llvm.store %99, %101 : i32, !llvm.ptr
%102 = arith.constant 0 : i32
%103 = llvm.mlir.constant(1 : i64) : i64
%104 = llvm.alloca %103 x i32 : (i64) -> !llvm.ptr
llvm.store %102, %104 : i32, !llvm.ptr
cf.br ^bb15
^bb15:
%105 = llvm.load %104 : !llvm.ptr -> i32
%106 = llvm.load %101 : !llvm.ptr -> i32
%107 = arith.cmpi slt, %105, %106 : i32
cf.cond_br %107, ^bb16, ^bb17
^bb16:
%109 = llvm.load %104 : !llvm.ptr -> i32
%110 = arith.extsi %109 : i32 to i64
%111 = llvm.getelementptr %60[%110] : (!llvm.ptr, i64) -> !llvm.ptr, i32
%108 = llvm.load %111 : !llvm.ptr -> i32
%112 = arith.constant 0 : i32
%113 = llvm.mlir.constant(1 : i64) : i64
%114 = llvm.alloca %113 x i32 : (i64) -> !llvm.ptr
llvm.store %112, %114 : i32, !llvm.ptr
cf.br ^bb18
^bb18:
%115 = llvm.load %114 : !llvm.ptr -> i32
%116 = arith.constant 16 : i32
%117 = arith.cmpi slt, %115, %116 : i32
cf.cond_br %117, ^bb19, ^bb20
^bb19:
%119 = llvm.load %114 : !llvm.ptr -> i32
%118 = func.call @update_reachable(%108, %119) : (i32, i32) -> i32
%121 = arith.extsi %118 : i32 to i64
%122 = llvm.getelementptr %65[%121] : (!llvm.ptr, i64) -> !llvm.ptr, i32
%120 = llvm.load %122 : !llvm.ptr -> i32
%123 = arith.constant 0 : i32
%124 = arith.cmpi slt, %120, %123 : i32
cf.cond_br %124, ^bb21, ^bb22
^bb21:
%125 = llvm.load %101 : !llvm.ptr -> i32
%126 = arith.extsi %118 : i32 to i64
%127 = llvm.getelementptr %65[%126] : (!llvm.ptr, i64) -> !llvm.ptr, i32
llvm.store %125, %127 : i32, !llvm.ptr
%128 = llvm.load %101 : !llvm.ptr -> i32
%129 = arith.extsi %128 : i32 to i64
%130 = llvm.getelementptr %60[%129] : (!llvm.ptr, i64) -> !llvm.ptr, i32
llvm.store %118, %130 : i32, !llvm.ptr
%131 = llvm.load %101 : !llvm.ptr -> i32
%132 = arith.constant 1 : i32
%133 = arith.addi %131, %132 : i32
llvm.store %133, %101 : i32, !llvm.ptr
cf.br ^bb23
^bb22:
cf.br ^bb23
^bb23:
%134 = llvm.load %114 : !llvm.ptr -> i32
%135 = arith.constant 1 : i32
%136 = arith.addi %134, %135 : i32
llvm.store %136, %114 : i32, !llvm.ptr
cf.br ^bb18
^bb20:
%137 = llvm.load %104 : !llvm.ptr -> i32
%138 = arith.constant 1 : i32
%139 = arith.addi %137, %138 : i32
llvm.store %139, %104 : i32, !llvm.ptr
cf.br ^bb15
^bb17:
%141 = llvm.load %101 : !llvm.ptr -> i32
%142 = arith.constant 16 : i32
%143 = arith.muli %141, %142 : i32
%144 = arith.extsi %143 : i32 to i64
%145 = arith.constant 4 : i32
%146 = arith.extsi %145 : i32 to i64
%140 = func.call @calloc(%144, %146) : (i64, i64) -> !llvm.ptr
%147 = llvm.mlir.zero : !llvm.ptr
%148 = llvm.icmp "eq" %140, %147 : !llvm.ptr
cf.cond_br %148, ^bb24, ^bb25
^bb24:
%149 = arith.constant 1 : i32
func.return %149 : i32
^bb25:
cf.br ^bb26
^bb26:
%150 = arith.constant 0 : i32
llvm.store %150, %79 : i32, !llvm.ptr
cf.br ^bb27
^bb27:
%151 = llvm.load %79 : !llvm.ptr -> i32
%152 = llvm.load %101 : !llvm.ptr -> i32
%153 = arith.cmpi slt, %151, %152 : i32
cf.cond_br %153, ^bb28, ^bb29
^bb28:
%154 = arith.constant 0 : i32
%155 = llvm.mlir.constant(1 : i64) : i64
%156 = llvm.alloca %155 x i32 : (i64) -> !llvm.ptr
llvm.store %154, %156 : i32, !llvm.ptr
cf.br ^bb30
^bb30:
%157 = llvm.load %156 : !llvm.ptr -> i32
%158 = arith.constant 16 : i32
%159 = arith.cmpi slt, %157, %158 : i32
cf.cond_br %159, ^bb31, ^bb32
^bb31:
%163 = llvm.load %79 : !llvm.ptr -> i32
%164 = arith.extsi %163 : i32 to i64
%165 = llvm.getelementptr %60[%164] : (!llvm.ptr, i64) -> !llvm.ptr, i32
%162 = llvm.load %165 : !llvm.ptr -> i32
%166 = llvm.load %156 : !llvm.ptr -> i32
%161 = func.call @update_reachable(%162, %166) : (i32, i32) -> i32
%167 = arith.extsi %161 : i32 to i64
%168 = llvm.getelementptr %65[%167] : (!llvm.ptr, i64) -> !llvm.ptr, i32
%160 = llvm.load %168 : !llvm.ptr -> i32
%169 = llvm.load %79 : !llvm.ptr -> i32
%170 = arith.constant 16 : i32
%171 = arith.muli %169, %170 : i32
%172 = llvm.load %156 : !llvm.ptr -> i32
%173 = arith.addi %171, %172 : i32
%174 = arith.extsi %173 : i32 to i64
%175 = llvm.getelementptr %140[%174] : (!llvm.ptr, i64) -> !llvm.ptr, i32
llvm.store %160, %175 : i32, !llvm.ptr
%176 = llvm.load %156 : !llvm.ptr -> i32
%177 = arith.constant 1 : i32
%178 = arith.addi %176, %177 : i32
llvm.store %178, %156 : i32, !llvm.ptr
cf.br ^bb30
^bb32:
%179 = llvm.load %79 : !llvm.ptr -> i32
%180 = arith.constant 1 : i32
%181 = arith.addi %179, %180 : i32
llvm.store %181, %79 : i32, !llvm.ptr
cf.br ^bb27
^bb29:
%183 = llvm.load %101 : !llvm.ptr -> i32
%184 = arith.constant 1 : i32
%185 = arith.addi %59, %184 : i32
%186 = arith.muli %183, %185 : i32
%187 = arith.extsi %186 : i32 to i64
%188 = arith.constant 8 : i32
%189 = arith.extsi %188 : i32 to i64
%182 = func.call @calloc(%187, %189) : (i64, i64) -> !llvm.ptr
%191 = llvm.load %101 : !llvm.ptr -> i32
%192 = arith.constant 1 : i32
%193 = arith.addi %59, %192 : i32
%194 = arith.muli %191, %193 : i32
%195 = arith.extsi %194 : i32 to i64
%196 = arith.constant 8 : i32
%197 = arith.extsi %196 : i32 to i64
%190 = func.call @calloc(%195, %197) : (i64, i64) -> !llvm.ptr
%198 = llvm.mlir.zero : !llvm.ptr
%199 = llvm.icmp "eq" %182, %198 : !llvm.ptr
%200 = scf.if %199 -> (i1) {
%201 = arith.constant true
scf.yield %201 : i1
} else {
%202 = llvm.mlir.zero : !llvm.ptr
%203 = llvm.icmp "eq" %190, %202 : !llvm.ptr
scf.yield %203 : i1
}
cf.cond_br %200, ^bb33, ^bb34
^bb33:
%204 = arith.constant 1 : i32
func.return %204 : i32
^bb34:
cf.br ^bb35
^bb35:
%205 = arith.constant 1 : i32
%206 = arith.constant 0 : i32
%207 = arith.constant 1 : i32
%208 = arith.addi %59, %207 : i32
%209 = arith.muli %206, %208 : i32
%210 = arith.constant 0 : i32
%211 = arith.addi %209, %210 : i32
%212 = arith.extsi %205 : i32 to i64
%213 = arith.extsi %211 : i32 to i64
%214 = llvm.getelementptr %182[%213] : (!llvm.ptr, i64) -> !llvm.ptr, i64
llvm.store %212, %214 : i64, !llvm.ptr
%215 = arith.constant 0 : i32
%216 = llvm.mlir.constant(1 : i64) : i64
%217 = llvm.alloca %216 x i32 : (i64) -> !llvm.ptr
llvm.store %215, %217 : i32, !llvm.ptr
cf.br ^bb36
^bb36:
%218 = llvm.load %217 : !llvm.ptr -> i32
%219 = arith.constant 13 : i32
%220 = arith.cmpi slt, %218, %219 : i32
cf.cond_br %220, ^bb37, ^bb38
^bb37:
%221 = arith.constant 0 : i32
%222 = arith.extsi %221 : i32 to i64
%223 = llvm.mlir.constant(1 : i64) : i64
%224 = llvm.alloca %223 x i64 : (i64) -> !llvm.ptr
llvm.store %222, %224 : i64, !llvm.ptr
cf.br ^bb39
^bb39:
%225 = llvm.load %224 : !llvm.ptr -> i64
%226 = llvm.load %101 : !llvm.ptr -> i32
%227 = arith.constant 1 : i32
%228 = arith.addi %59, %227 : i32
%229 = arith.muli %226, %228 : i32
%230 = arith.extsi %229 : i32 to i64
%231 = arith.cmpi slt, %225, %230 : i64
cf.cond_br %231, ^bb40, ^bb41
^bb40:
%232 = arith.constant 0 : i32
%233 = llvm.load %224 : !llvm.ptr -> i64
%234 = arith.extsi %232 : i32 to i64
%235 = llvm.getelementptr %190[%233] : (!llvm.ptr, i64) -> !llvm.ptr, i64
llvm.store %234, %235 : i64, !llvm.ptr
%236 = llvm.load %224 : !llvm.ptr -> i64
%237 = arith.constant 1 : i32
%239 = arith.extsi %237 : i32 to i64
%238 = arith.addi %236, %239 : i64
llvm.store %238, %224 : i64, !llvm.ptr
cf.br ^bb39
^bb41:
%240 = arith.constant 0 : i32
llvm.store %240, %79 : i32, !llvm.ptr
cf.br ^bb42
^bb42:
%241 = llvm.load %79 : !llvm.ptr -> i32
%242 = llvm.load %101 : !llvm.ptr -> i32
%243 = arith.cmpi slt, %241, %242 : i32
cf.cond_br %243, ^bb43, ^bb44
^bb43:
%244 = arith.constant 0 : i32
%245 = llvm.mlir.constant(1 : i64) : i64
%246 = llvm.alloca %245 x i32 : (i64) -> !llvm.ptr
llvm.store %244, %246 : i32, !llvm.ptr
%247 = llvm.load %79 : !llvm.ptr -> i32
%248 = arith.constant 1 : i32
%249 = arith.addi %59, %248 : i32
%250 = arith.muli %247, %249 : i32
%251 = arith.constant 0 : i32
%252 = llvm.mlir.constant(1 : i64) : i64
%253 = llvm.alloca %252 x i32 : (i64) -> !llvm.ptr
llvm.store %251, %253 : i32, !llvm.ptr
cf.br ^bb45
^bb45:
%254 = llvm.load %246 : !llvm.ptr -> i32
%255 = arith.cmpi sle, %254, %59 : i32
cf.cond_br %255, ^bb46, ^bb47
^bb46:
%257 = llvm.load %246 : !llvm.ptr -> i32
%258 = arith.addi %250, %257 : i32
%259 = arith.extsi %258 : i32 to i64
%260 = llvm.getelementptr %182[%259] : (!llvm.ptr, i64) -> !llvm.ptr, i64
%256 = llvm.load %260 : !llvm.ptr -> i64
%261 = arith.constant 0 : i32
%263 = arith.extsi %261 : i32 to i64
%262 = arith.cmpi ne, %256, %263 : i64
cf.cond_br %262, ^bb48, ^bb49
^bb48:
%264 = arith.constant 1 : i32
llvm.store %264, %253 : i32, !llvm.ptr
cf.br ^bb47
^bb49:
cf.br ^bb50
^bb50:
%265 = llvm.load %246 : !llvm.ptr -> i32
%266 = arith.constant 1 : i32
%267 = arith.addi %265, %266 : i32
llvm.store %267, %246 : i32, !llvm.ptr
cf.br ^bb45
^bb47:
%268 = llvm.load %253 : !llvm.ptr -> i32
%269 = arith.constant 1 : i32
%270 = arith.cmpi eq, %268, %269 : i32
cf.cond_br %270, ^bb51, ^bb52
^bb51:
%271 = arith.constant 0 : i32
%272 = llvm.mlir.constant(1 : i64) : i64
%273 = llvm.alloca %272 x i32 : (i64) -> !llvm.ptr
llvm.store %271, %273 : i32, !llvm.ptr
cf.br ^bb54
^bb54:
%274 = llvm.load %273 : !llvm.ptr -> i32
%275 = arith.constant 16 : i32
%276 = arith.cmpi slt, %274, %275 : i32
cf.cond_br %276, ^bb55, ^bb56
^bb55:
%278 = llvm.load %79 : !llvm.ptr -> i32
%279 = arith.constant 16 : i32
%280 = arith.muli %278, %279 : i32
%281 = llvm.load %273 : !llvm.ptr -> i32
%282 = arith.addi %280, %281 : i32
%283 = arith.extsi %282 : i32 to i64
%284 = llvm.getelementptr %140[%283] : (!llvm.ptr, i64) -> !llvm.ptr, i32
%277 = llvm.load %284 : !llvm.ptr -> i32
%286 = llvm.load %273 : !llvm.ptr -> i32
%285 = func.call @popcount4(%286) : (i32) -> i32
%287 = arith.constant 0 : i32
llvm.store %287, %246 : i32, !llvm.ptr
cf.br ^bb57
^bb57:
%288 = llvm.load %246 : !llvm.ptr -> i32
%289 = arith.cmpi sle, %288, %59 : i32
cf.cond_br %289, ^bb58, ^bb59
^bb58:
%291 = llvm.load %246 : !llvm.ptr -> i32
%292 = arith.addi %250, %291 : i32
%293 = arith.extsi %292 : i32 to i64
%294 = llvm.getelementptr %182[%293] : (!llvm.ptr, i64) -> !llvm.ptr, i64
%290 = llvm.load %294 : !llvm.ptr -> i64
%295 = llvm.load %246 : !llvm.ptr -> i32
%296 = arith.addi %295, %285 : i32
%297 = arith.constant 0 : i32
%299 = arith.extsi %297 : i32 to i64
%298 = arith.cmpi ne, %290, %299 : i64
%300 = scf.if %298 -> (i1) {
%301 = arith.cmpi sle, %296, %59 : i32
scf.yield %301 : i1
} else {
%302 = arith.constant false
scf.yield %302 : i1
}
cf.cond_br %300, ^bb60, ^bb61
^bb60:
%304 = arith.constant 1 : i32
%305 = arith.addi %59, %304 : i32
%306 = arith.muli %277, %305 : i32
%307 = arith.addi %306, %296 : i32
%308 = arith.extsi %307 : i32 to i64
%309 = llvm.getelementptr %190[%308] : (!llvm.ptr, i64) -> !llvm.ptr, i64
%303 = llvm.load %309 : !llvm.ptr -> i64
%310 = arith.addi %303, %290 : i64
%311 = arith.constant 1 : i32
%312 = arith.addi %59, %311 : i32
%313 = arith.muli %277, %312 : i32
%314 = arith.addi %313, %296 : i32
%315 = arith.extsi %314 : i32 to i64
%316 = llvm.getelementptr %190[%315] : (!llvm.ptr, i64) -> !llvm.ptr, i64
llvm.store %310, %316 : i64, !llvm.ptr
cf.br ^bb62
^bb61:
cf.br ^bb62
^bb62:
%317 = llvm.load %246 : !llvm.ptr -> i32
%318 = arith.constant 1 : i32
%319 = arith.addi %317, %318 : i32
llvm.store %319, %246 : i32, !llvm.ptr
cf.br ^bb57
^bb59:
%320 = llvm.load %273 : !llvm.ptr -> i32
%321 = arith.constant 1 : i32
%322 = arith.addi %320, %321 : i32
llvm.store %322, %273 : i32, !llvm.ptr
cf.br ^bb54
^bb56:
cf.br ^bb53
^bb52:
cf.br ^bb53
^bb53:
%323 = llvm.load %79 : !llvm.ptr -> i32
%324 = arith.constant 1 : i32
%325 = arith.addi %323, %324 : i32
llvm.store %325, %79 : i32, !llvm.ptr
cf.br ^bb42
^bb44:
%326 = arith.constant 0 : i32
%327 = arith.extsi %326 : i32 to i64
llvm.store %327, %224 : i64, !llvm.ptr
cf.br ^bb63
^bb63:
%328 = llvm.load %224 : !llvm.ptr -> i64
%329 = llvm.load %101 : !llvm.ptr -> i32
%330 = arith.constant 1 : i32
%331 = arith.addi %59, %330 : i32
%332 = arith.muli %329, %331 : i32
%333 = arith.extsi %332 : i32 to i64
%334 = arith.cmpi slt, %328, %333 : i64
cf.cond_br %334, ^bb64, ^bb65
^bb64:
%336 = llvm.load %224 : !llvm.ptr -> i64
%337 = llvm.getelementptr %190[%336] : (!llvm.ptr, i64) -> !llvm.ptr, i64
%335 = llvm.load %337 : !llvm.ptr -> i64
%338 = llvm.load %224 : !llvm.ptr -> i64
%339 = llvm.getelementptr %182[%338] : (!llvm.ptr, i64) -> !llvm.ptr, i64
llvm.store %335, %339 : i64, !llvm.ptr
%340 = llvm.load %224 : !llvm.ptr -> i64
%341 = arith.constant 1 : i32
%343 = arith.extsi %341 : i32 to i64
%342 = arith.addi %340, %343 : i64
llvm.store %342, %224 : i64, !llvm.ptr
cf.br ^bb63
^bb65:
%344 = llvm.load %217 : !llvm.ptr -> i32
%345 = arith.constant 1 : i32
%346 = arith.addi %344, %345 : i32
llvm.store %346, %217 : i32, !llvm.ptr
cf.br ^bb36
^bb38:
%347 = arith.constant 1 : i32
%348 = arith.constant 15 : i32
%349 = arith.shli %347, %348 : i32
%350 = arith.constant 0 : i32
%351 = arith.extsi %350 : i32 to i64
%352 = llvm.mlir.constant(1 : i64) : i64
%353 = llvm.alloca %352 x i64 : (i64) -> !llvm.ptr
llvm.store %351, %353 : i64, !llvm.ptr
%354 = arith.constant 0 : i32
llvm.store %354, %79 : i32, !llvm.ptr
cf.br ^bb66
^bb66:
%355 = llvm.load %79 : !llvm.ptr -> i32
%356 = llvm.load %101 : !llvm.ptr -> i32
%357 = arith.cmpi slt, %355, %356 : i32
cf.cond_br %357, ^bb67, ^bb68
^bb67:
%359 = llvm.load %79 : !llvm.ptr -> i32
%360 = arith.extsi %359 : i32 to i64
%361 = llvm.getelementptr %60[%360] : (!llvm.ptr, i64) -> !llvm.ptr, i32
%358 = llvm.load %361 : !llvm.ptr -> i32
%362 = arith.andi %358, %349 : i32
%363 = arith.constant 0 : i32
%364 = arith.cmpi ne, %362, %363 : i32
cf.cond_br %364, ^bb69, ^bb70
^bb69:
%365 = arith.constant 4 : i32
%366 = llvm.mlir.constant(1 : i64) : i64
%367 = llvm.alloca %366 x i32 : (i64) -> !llvm.ptr
llvm.store %365, %367 : i32, !llvm.ptr
cf.br ^bb72
^bb72:
%368 = llvm.load %367 : !llvm.ptr -> i32
%369 = arith.cmpi sle, %368, %59 : i32
cf.cond_br %369, ^bb73, ^bb74
^bb73:
%370 = llvm.load %353 : !llvm.ptr -> i64
%372 = llvm.load %79 : !llvm.ptr -> i32
%373 = arith.constant 1 : i32
%374 = arith.addi %59, %373 : i32
%375 = arith.muli %372, %374 : i32
%376 = llvm.load %367 : !llvm.ptr -> i32
%377 = arith.addi %375, %376 : i32
%378 = arith.extsi %377 : i32 to i64
%379 = llvm.getelementptr %182[%378] : (!llvm.ptr, i64) -> !llvm.ptr, i64
%371 = llvm.load %379 : !llvm.ptr -> i64
%380 = arith.addi %370, %371 : i64
llvm.store %380, %353 : i64, !llvm.ptr
%381 = llvm.load %367 : !llvm.ptr -> i32
%382 = arith.constant 1 : i32
%383 = arith.addi %381, %382 : i32
llvm.store %383, %367 : i32, !llvm.ptr
cf.br ^bb72
^bb74:
cf.br ^bb71
^bb70:
cf.br ^bb71
^bb71:
%384 = llvm.load %79 : !llvm.ptr -> i32
%385 = arith.constant 1 : i32
%386 = arith.addi %384, %385 : i32
llvm.store %386, %79 : i32, !llvm.ptr
cf.br ^bb66
^bb68:
%387 = llvm.mlir.addressof @str_0 : !llvm.ptr
%388 = llvm.load %353 : !llvm.ptr -> i64
%389 = llvm.call @printf(%387, %388) vararg(!llvm.func<i32 (ptr, ...)>) : (!llvm.ptr, i64) -> i32
func.call @free(%190) : (!llvm.ptr) -> ()
func.call @free(%182) : (!llvm.ptr) -> ()
func.call @free(%140) : (!llvm.ptr) -> ()
func.call @free(%65) : (!llvm.ptr) -> ()
func.call @free(%60) : (!llvm.ptr) -> ()
%395 = arith.constant 0 : i32
func.return %395 : i32
}
}