Implemented exhaustive testing of n nested powerlosses

As expected this takes a significant amount of time (~10 minutes for all
1 powerlosses, >10 hours for all 2 powerlosses) but this may be reducible in
the future by optimizing tests for powerloss testing. Currently
test_files does a lot of work that doesn't really have testing value.

bd/lfs_testbd.c

@@ -29,68 +29,65 @@
 // Note we can only modify a block if we have exclusive access to it (rc == 1)
 //
 
-// TODO
-__attribute__((unused))
-static void lfs_testbd_incblock(lfs_testbd_t *bd, lfs_block_t block) {
-    if (bd->blocks[block]) {
-        bd->blocks[block]->rc += 1;
+static lfs_testbd_block_t *lfs_testbd_incblock(lfs_testbd_block_t *block) {
+    if (block) {
+        block->rc += 1;
     }
+    return block;
 }
 
-static void lfs_testbd_decblock(lfs_testbd_t *bd, lfs_block_t block) {
-    if (bd->blocks[block]) {
-        bd->blocks[block]->rc -= 1;
-        if (bd->blocks[block]->rc == 0) {
-            free(bd->blocks[block]);
-            bd->blocks[block] = NULL;
+static void lfs_testbd_decblock(lfs_testbd_block_t *block) {
+    if (block) {
+        block->rc -= 1;
+        if (block->rc == 0) {
+            free(block);
         }
     }
 }
 
-static const lfs_testbd_block_t *lfs_testbd_getblock(lfs_testbd_t *bd,
-        lfs_block_t block) {
-    return bd->blocks[block];
-}
-
-static lfs_testbd_block_t *lfs_testbd_mutblock(lfs_testbd_t *bd,
-        lfs_block_t block, lfs_size_t block_size) {
-    if (bd->blocks[block] && bd->blocks[block]->rc == 1) {
+static lfs_testbd_block_t *lfs_testbd_mutblock(
+        const struct lfs_config *cfg,
+        lfs_testbd_block_t **block) {
+    lfs_testbd_block_t *block_ = *block;
+    if (block_ && block_->rc == 1) {
         // rc == 1? can modify
-        return bd->blocks[block];
+        return block_;
 
-    } else if (bd->blocks[block]) {
+    } else if (block_) {
         // rc > 1? need to create a copy
-        lfs_testbd_block_t *b = malloc(
-                sizeof(lfs_testbd_block_t) + block_size);
-        if (!b) {
+        lfs_testbd_block_t *nblock = malloc(
+                sizeof(lfs_testbd_block_t) + cfg->block_size);
+        if (!nblock) {
             return NULL;
         }
 
-        memcpy(b, bd->blocks[block], sizeof(lfs_testbd_block_t) + block_size);
-        b->rc = 1;
+        memcpy(nblock, block_,
+                sizeof(lfs_testbd_block_t) + cfg->block_size);
+        nblock->rc = 1;
 
-        lfs_testbd_decblock(bd, block);
-        bd->blocks[block] = b;
-        return b;
+        lfs_testbd_decblock(block_);
+        *block = nblock;
+        return nblock;
 
     } else {
         // no block? need to allocate
-        lfs_testbd_block_t *b = malloc(
-                sizeof(lfs_testbd_block_t) + block_size);
-        if (!b) {
+        lfs_testbd_block_t *nblock = malloc(
+                sizeof(lfs_testbd_block_t) + cfg->block_size);
+        if (!nblock) {
             return NULL;
         }
 
-        b->rc = 1;
-        b->wear = 0;
+        nblock->rc = 1;
+        nblock->wear = 0;
 
         // zero for consistency
-        memset(b->data,
+        lfs_testbd_t *bd = cfg->context;
+        memset(nblock->data,
                 (bd->cfg->erase_value != -1) ? bd->cfg->erase_value : 0,
-                block_size);
+                cfg->block_size);
 
-        bd->blocks[block] = b;
-        return b;
+        *block = nblock;
+        return nblock;
     }
 }
 
@@ -129,22 +126,25 @@ int lfs_testbd_createcfg(const struct lfs_config *cfg, const char *path,
 
     // setup testing things
     bd->power_cycles = bd->cfg->power_cycles;
-    bd->disk_fd = -1;
-    bd->disk_scratch_block = NULL;
-
-    bd->branches = NULL;
-    bd->branch_capacity = 0;
-    bd->branch_count = 0;
+    bd->disk = NULL;
 
     if (bd->cfg->disk_path) {
+        bd->disk = malloc(sizeof(lfs_testbd_disk_t));
+        if (!bd->disk) {
+            LFS_TESTBD_TRACE("lfs_testbd_createcfg -> %d", LFS_ERR_NOMEM);
+            return LFS_ERR_NOMEM;
+        }
+        bd->disk->rc = 1;
+        bd->disk->scratch = NULL;
+
         #ifdef _WIN32
-        bd->disk_fd = open(bd->cfg->disk_path,
+        bd->disk->fd = open(bd->cfg->disk_path,
                 O_RDWR | O_CREAT | O_BINARY, 0666);
         #else
-        bd->disk_fd = open(bd->cfg->disk_path,
+        bd->disk->fd = open(bd->cfg->disk_path,
                 O_RDWR | O_CREAT, 0666);
         #endif
-        if (bd->disk_fd < 0) {
+        if (bd->disk->fd < 0) {
             int err = -errno;
             LFS_TESTBD_TRACE("lfs_testbd_create -> %d", err);
             return err;
@@ -153,12 +153,12 @@ int lfs_testbd_createcfg(const struct lfs_config *cfg, const char *path,
         // if we're emulating erase values, we can keep a block around in
         // memory of just the erase state to speed up emulated erases
         if (bd->cfg->erase_value != -1) {
-            bd->disk_scratch_block = malloc(cfg->block_size);
-            if (!bd->disk_scratch_block) {
+            bd->disk->scratch = malloc(cfg->block_size);
+            if (!bd->disk->scratch) {
                 LFS_TESTBD_TRACE("lfs_testbd_createcfg -> %d", LFS_ERR_NOMEM);
                 return LFS_ERR_NOMEM;
             }
-            memset(bd->disk_scratch_block,
+            memset(bd->disk->scratch,
                     bd->cfg->erase_value,
                     cfg->block_size);
         }
@@ -191,16 +191,18 @@ int lfs_testbd_destroy(const struct lfs_config *cfg) {
 
     // decrement reference counts
     for (lfs_block_t i = 0; i < cfg->block_count; i++) {
-        lfs_testbd_decblock(bd, i);
+        lfs_testbd_decblock(bd->blocks[i]);
     }
-
-    // free memory
     free(bd->blocks);
-    free(bd->branches);
 
-    if (bd->disk_fd >= 0) {
-        close(bd->disk_fd);
-        free(bd->disk_scratch_block);
+    // clean up other resources 
+    if (bd->disk) {
+        bd->disk->rc -= 1;
+        if (bd->disk->rc == 0) {
+            close(bd->disk->fd);
+            free(bd->disk->scratch);
+            free(bd->disk);
+        }
     }
 
     LFS_TESTBD_TRACE("lfs_testbd_destroy -> %d", 0);
@@ -225,7 +227,7 @@ int lfs_testbd_read(const struct lfs_config *cfg, lfs_block_t block,
     LFS_ASSERT(off+size <= cfg->block_size);
 
     // get the block
-    const lfs_testbd_block_t *b = lfs_testbd_getblock(bd, block);
+    const lfs_testbd_block_t *b = bd->blocks[block];
     if (b) {
         // block bad?
         if (bd->cfg->erase_cycles && b->wear >= bd->cfg->erase_cycles &&
@@ -273,7 +275,7 @@ int lfs_testbd_prog(const struct lfs_config *cfg, lfs_block_t block,
     LFS_ASSERT(off+size <= cfg->block_size);
 
     // get the block
-    lfs_testbd_block_t *b = lfs_testbd_mutblock(bd, block, cfg->block_size);
+    lfs_testbd_block_t *b = lfs_testbd_mutblock(cfg, &bd->blocks[block]);
     if (!b) {
         LFS_TESTBD_TRACE("lfs_testbd_prog -> %d", LFS_ERR_NOMEM);
         return LFS_ERR_NOMEM;
@@ -305,8 +307,8 @@ int lfs_testbd_prog(const struct lfs_config *cfg, lfs_block_t block,
     memcpy(&b->data[off], buffer, size);
 
     // mirror to disk file?
-    if (bd->disk_fd >= 0) {
-        off_t res1 = lseek(bd->disk_fd,
+    if (bd->disk) {
+        off_t res1 = lseek(bd->disk->fd,
                 (off_t)block*cfg->block_size + (off_t)off,
                 SEEK_SET);
         if (res1 < 0) {
@@ -315,7 +317,7 @@ int lfs_testbd_prog(const struct lfs_config *cfg, lfs_block_t block,
             return err;
         }
 
-        ssize_t res2 = write(bd->disk_fd, buffer, size);
+        ssize_t res2 = write(bd->disk->fd, buffer, size);
         if (res2 < 0) {
             int err = -errno;
             LFS_TESTBD_TRACE("lfs_testbd_prog -> %d", err);
@@ -344,15 +346,6 @@ int lfs_testbd_prog(const struct lfs_config *cfg, lfs_block_t block,
         }
     }
 
-//    // track power-loss branch?
-//    if (bd->cfg->track_branches) {
-//        int err = lfs_testbd_trackbranch(bd);
-//        if (err) {
-//            LFS_TESTBD_TRACE("lfs_testbd_prog -> %d", err);
-//            return err;
-//        }
-//    }
-
     LFS_TESTBD_TRACE("lfs_testbd_prog -> %d", 0);
     return 0;
 }
@@ -365,7 +358,7 @@ int lfs_testbd_erase(const struct lfs_config *cfg, lfs_block_t block) {
     LFS_ASSERT(block < cfg->block_count);
 
     // get the block
-    lfs_testbd_block_t *b = lfs_testbd_mutblock(bd, block, cfg->block_size);
+    lfs_testbd_block_t *b = lfs_testbd_mutblock(cfg, &bd->blocks[block]);
     if (!b) {
         LFS_TESTBD_TRACE("lfs_testbd_prog -> %d", LFS_ERR_NOMEM);
         return LFS_ERR_NOMEM;
@@ -394,8 +387,8 @@ int lfs_testbd_erase(const struct lfs_config *cfg, lfs_block_t block) {
         memset(b->data, bd->cfg->erase_value, cfg->block_size);
 
         // mirror to disk file?
-        if (bd->disk_fd >= 0) {
-            off_t res1 = lseek(bd->disk_fd,
+        if (bd->disk) {
+            off_t res1 = lseek(bd->disk->fd,
                     (off_t)block*cfg->block_size,
                     SEEK_SET);
             if (res1 < 0) {
@@ -404,8 +397,8 @@ int lfs_testbd_erase(const struct lfs_config *cfg, lfs_block_t block) {
                 return err;
             }
 
-            ssize_t res2 = write(bd->disk_fd,
-                    bd->disk_scratch_block,
+            ssize_t res2 = write(bd->disk->fd,
+                    bd->disk->scratch,
                     cfg->block_size);
             if (res2 < 0) {
                 int err = -errno;
@@ -436,15 +429,6 @@ int lfs_testbd_erase(const struct lfs_config *cfg, lfs_block_t block) {
         }
     }
 
-//    // track power-loss branch?
-//    if (bd->cfg->track_branches) {
-//        int err = lfs_testbd_trackbranch(bd);
-//        if (err) {
-//            LFS_TESTBD_TRACE("lfs_testbd_prog -> %d", err);
-//            return err;
-//        }
-//    }
-
     LFS_TESTBD_TRACE("lfs_testbd_prog -> %d", 0);
     return 0;
 }
@@ -472,7 +456,7 @@ lfs_testbd_swear_t lfs_testbd_getwear(const struct lfs_config *cfg,
 
     // get the wear
     lfs_testbd_wear_t wear;
-    const lfs_testbd_block_t *b = lfs_testbd_getblock(bd, block);
+    const lfs_testbd_block_t *b = bd->blocks[block];
     if (b) {
         wear = b->wear;
     } else {
@@ -492,7 +476,7 @@ int lfs_testbd_setwear(const struct lfs_config *cfg,
     LFS_ASSERT(block < cfg->block_count);
 
     // set the wear
-    lfs_testbd_block_t *b = lfs_testbd_mutblock(bd, block, cfg->block_size);
+    lfs_testbd_block_t *b = lfs_testbd_mutblock(cfg, &bd->blocks[block]);
     if (!b) {
         LFS_TESTBD_TRACE("lfs_testbd_setwear -> %"PRIu32, LFS_ERR_NOMEM);
         return LFS_ERR_NOMEM;
@@ -524,23 +508,30 @@ int lfs_testbd_setpowercycles(const struct lfs_config *cfg,
     return 0;
 }
 
-//int lfs_testbd_getbranch(const struct lfs_config *cfg,
-//        lfs_testbd_powercycles_t branch, lfs_testbd_t *bd) {
-//    LFS_TESTBD_TRACE("lfs_testbd_getbranch(%p, %zu, %p)",
-//            (void*)cfg, branch, bd);
-//    lfs_testbd_t *bd = cfg->context;
-//
-//    // TODO
-//
-//    LFS_TESTBD_TRACE("lfs_testbd_getbranch -> %d", 0);
-//    return 0;
-//}
-
-lfs_testbd_spowercycles_t lfs_testbd_getbranchcount(
-        const struct lfs_config *cfg) {
-    LFS_TESTBD_TRACE("lfs_testbd_getbranchcount(%p)", (void*)cfg);
+int lfs_testbd_copy(const struct lfs_config *cfg, lfs_testbd_t *copy) {
+    LFS_TESTBD_TRACE("lfs_testbd_copy(%p, %p)", (void*)cfg, (void*)copy);
     lfs_testbd_t *bd = cfg->context;
 
-    LFS_TESTBD_TRACE("lfs_testbd_getbranchcount -> %"PRIu32, bd->branch_count);
-    return bd->branch_count;
+    // lazily copy over our block array
+    copy->blocks = malloc(cfg->block_count * sizeof(lfs_testbd_block_t*));
+    if (!copy->blocks) {
+        LFS_TESTBD_TRACE("lfs_testbd_copy -> %d", LFS_ERR_NOMEM);
+        return LFS_ERR_NOMEM;
+    }
+
+    for (size_t i = 0; i < cfg->block_count; i++) {
+        copy->blocks[i] = lfs_testbd_incblock(bd->blocks[i]);
+    }
+
+    // other state
+    copy->power_cycles = bd->power_cycles;
+    copy->disk = bd->disk;
+    if (copy->disk) {
+        copy->disk->rc += 1;
+    }
+    copy->cfg = bd->cfg;
+
+    LFS_TESTBD_TRACE("lfs_testbd_copy -> %d", 0);
+    return 0;
 }
+

bd/lfs_testbd.h

@@ -119,6 +119,13 @@ typedef struct lfs_testbd_block {
     uint8_t data[];
 } lfs_testbd_block_t;
 
+// Disk mirror
+typedef struct lfs_testbd_disk {
+    uint32_t rc;
+    int fd;
+    uint8_t *scratch;
+} lfs_testbd_disk_t;
+
 // testbd state
 typedef struct lfs_testbd {
     // array of copy-on-write blocks
@@ -126,31 +133,7 @@ typedef struct lfs_testbd {
 
     // some other test state
     uint32_t power_cycles;
-    int disk_fd;
-    uint8_t *disk_scratch_block;
-
-    // array of tracked branches
-    struct lfs_testbd *branches;
-    lfs_testbd_powercycles_t branch_count;
-    lfs_testbd_powercycles_t branch_capacity;
-
-    // TODO file?
-    
-
-//    union {
-//        struct {
-//            lfs_filebd_t bd;
-//        } file;
-//        struct {
-//            lfs_rambd_t bd;
-//            struct lfs_rambd_config cfg;
-//        } ram;
-//    } u;
-//
-//    bool persist;
-//    uint32_t power_cycles;
-//    lfs_testbd_wear_t *wear;
-//    uint8_t *scratch;
+    lfs_testbd_disk_t *disk;
 
     const struct lfs_testbd_config *cfg;
 } lfs_testbd_t;
@@ -207,13 +190,8 @@ lfs_testbd_spowercycles_t lfs_testbd_getpowercycles(
 int lfs_testbd_setpowercycles(const struct lfs_config *cfg,
         lfs_testbd_powercycles_t power_cycles);
 
-// Get a power-loss branch, requires track_branches=true
-int lfs_testbd_getbranch(const struct lfs_config *cfg,
-        lfs_testbd_powercycles_t branch, lfs_testbd_t *bd);
-
-// Get the current number of power-loss branches
-lfs_testbd_spowercycles_t lfs_testbd_getbranchcount(
-        const struct lfs_config *cfg);
+// Create a copy-on-write copy of the state of this block device
+int lfs_testbd_copy(const struct lfs_config *cfg, lfs_testbd_t *copy);
 
 
 #ifdef __cplusplus

runners/test_runner.c

@@ -700,6 +700,7 @@ static void run_powerloss_linear(
 
     while (true) {
         if (!setjmp(powerloss_jmp)) {
+            // run the test
             case_->run(&cfg);
             break;
         }
@@ -780,6 +781,7 @@ static void run_powerloss_exponential(
 
     while (true) {
         if (!setjmp(powerloss_jmp)) {
+            // run the test
             case_->run(&cfg);
             break;
         }
@@ -858,6 +860,7 @@ static void run_powerloss_cycles(
 
     while (true) {
         if (!setjmp(powerloss_jmp)) {
+            // run the test
             case_->run(&cfg);
             break;
         }
@@ -883,15 +886,177 @@ static void run_powerloss_cycles(
     }
 }
 
-//static void run_powerloss_n(void *data,
-//
-//static void run_powerloss_incremental(void *data,
+struct powerloss_exhaustive_state {
+    struct lfs_config *cfg;
+
+    lfs_testbd_t *branches;
+    size_t branch_count;
+    size_t branch_capacity;
+};
+
+struct powerloss_exhaustive_cycles {
+    lfs_testbd_powercycles_t *cycles;
+    size_t cycle_count;
+    size_t cycle_capacity;
+};
+
+static void powerloss_exhaustive_branch(void *c) {
+    // append to branches
+    struct powerloss_exhaustive_state *state = c;
+    state->branch_count += 1;
+    if (state->branch_count > state->branch_capacity) {
+        state->branch_capacity = (2*state->branch_capacity > 4)
+                ? 2*state->branch_capacity
+                : 4;
+        state->branches = realloc(state->branches,
+                state->branch_capacity * sizeof(lfs_testbd_t));
+        if (!state->branches) {
+            fprintf(stderr, "error: exhaustive: out of memory\n");
+            exit(-1);
+        }
+    }
+
+    // create copy-on-write copy
+    int err = lfs_testbd_copy(state->cfg,
+            &state->branches[state->branch_count-1]);
+    if (err) {
+        fprintf(stderr, "error: exhaustive: could not create bd copy\n");
+        exit(-1);
+    }
+
+    // also trigger on next power cycle
+    lfs_testbd_setpowercycles(state->cfg, 1);
+}
+
+static void run_powerloss_exhaustive_layer(
+        const struct test_suite *suite,
+        const struct test_case *case_,
+        size_t perm,
+        struct lfs_config *cfg,
+        struct lfs_testbd_config *bdcfg,
+        size_t depth,
+        struct powerloss_exhaustive_cycles *cycles) {
+    (void)suite;
+
+    struct powerloss_exhaustive_state state = {
+        .cfg = cfg,
+        .branches = NULL,
+        .branch_count = 0,
+        .branch_capacity = 0,
+    };
+
+    // run through the test without additional powerlosses, collecting possible
+    // branches as we do so
+    lfs_testbd_setpowercycles(state.cfg, depth > 0 ? 1 : 0);
+    bdcfg->powerloss_data = &state;
+
+    // run the tests
+    case_->run(cfg);
+
+    // aggressively clean up memory here to try to keep our memory usage low
+    int err = lfs_testbd_destroy(cfg);
+    if (err) {
+        fprintf(stderr, "error: could not destroy block device: %d\n", err);
+        exit(-1);
+    }
+
+    // recurse into each branch
+    for (size_t i = 0; i < state.branch_count; i++) {
+        // first push and print the branch
+        cycles->cycle_count += 1;
+        if (cycles->cycle_count > cycles->cycle_capacity) {
+            cycles->cycle_capacity = (2*cycles->cycle_capacity > 4)
+                    ? 2*cycles->cycle_capacity
+                    : 4;
+            cycles->cycles = realloc(cycles->cycles,
+                    cycles->cycle_capacity * sizeof(lfs_testbd_powercycles_t));
+            if (!cycles->cycles) {
+                fprintf(stderr, "error: exhaustive: out of memory\n");
+                exit(-1);
+            }
+        }
+        cycles->cycles[cycles->cycle_count-1] = i;
+
+        printf("powerloss %s#%zu#", case_->id, perm);
+        leb16_print(cycles->cycles, cycles->cycle_count);
+        printf("\n");
+
+        // now recurse
+        cfg->context = &state.branches[i];
+        run_powerloss_exhaustive_layer(suite, case_, perm,
+                cfg, bdcfg, depth-1, cycles);
+
+        // pop the cycle
+        cycles->cycle_count -= 1;
+    }
+
+    // clean up memory
+    free(state.branches);
+}
+
+static void run_powerloss_exhaustive(
+        const struct test_suite *suite,
+        const struct test_case *case_,
+        size_t perm,
+        const lfs_testbd_powercycles_t *cycles,
+        size_t cycle_count) {
+    (void)cycles;
+    (void)suite;
+
+    // create block device and configuration
+    lfs_testbd_t bd;
+
+    struct lfs_config cfg = {
+        .context            = &bd,
+        .read               = lfs_testbd_read,
+        .prog               = lfs_testbd_prog,
+        .erase              = lfs_testbd_erase,
+        .sync               = lfs_testbd_sync,
+        .read_size          = READ_SIZE,
+        .prog_size          = PROG_SIZE,
+        .block_size         = BLOCK_SIZE,
+        .block_count        = BLOCK_COUNT,
+        .block_cycles       = BLOCK_CYCLES,
+        .cache_size         = CACHE_SIZE,
+        .lookahead_size     = LOOKAHEAD_SIZE,
+    };
+
+    struct lfs_testbd_config bdcfg = {
+        .erase_value        = ERASE_VALUE,
+        .erase_cycles       = ERASE_CYCLES,
+        .badblock_behavior  = BADBLOCK_BEHAVIOR,
+        .disk_path          = test_disk,
+        .read_delay         = test_read_delay,
+        .prog_delay         = test_prog_delay,
+        .erase_delay        = test_erase_delay,
+        .powerloss_behavior = POWERLOSS_BEHAVIOR,
+        .powerloss_cb       = powerloss_exhaustive_branch,
+        .powerloss_data     = NULL,
+    };
+
+    int err = lfs_testbd_createcfg(&cfg, test_disk, &bdcfg);
+    if (err) {
+        fprintf(stderr, "error: could not create block device: %d\n", err);
+        exit(-1);
+    }
+
+    // run the test, increasing power-cycles as power-loss events occur
+    printf("running %s#%zu\n", case_->id, perm);
+
+    // recursively exhaust each layer of powerlosses
+    run_powerloss_exhaustive_layer(suite, case_, perm,
+            &cfg, &bdcfg, cycle_count,
+            &(struct powerloss_exhaustive_cycles){NULL, 0, 0});
+
+    printf("finished %s#%zu\n", case_->id, perm);
+}
+
 
 const test_powerloss_t builtin_powerlosses[] = {
     {'0', "none",        run_powerloss_none,        NULL, 0},
     {'e', "exponential", run_powerloss_exponential, NULL, 0},
     {'l', "linear",      run_powerloss_linear,      NULL, 0},
-    //{'x', "exhaustive", run_powerloss_exhaustive}
+    {'x', "exhaustive",  run_powerloss_exhaustive,  NULL, SIZE_MAX},
     {0, NULL, NULL, NULL, 0},
 };
 
@@ -899,7 +1064,7 @@ const char *const builtin_powerlosses_help[] = {
     "Run with no power-losses.",
     "Run with linearly-decreasing power-losses.",
     "Run with exponentially-decreasing power-losses.",
-    //"Run a all permutations of power-losses, this may take a while.",
+    "Run a all permutations of power-losses, this may take a while.",
     "Run a all permutations of n power-losses.",
     "Run a custom comma-separated set of power-losses.",
     "Run a custom leb16-encoded set of power-losses.",
@@ -1273,9 +1438,6 @@ invalid_define:
                         }
                     }
 
-                    // exhaustive permutations
-                    // TODO
-
                     // comma-separated permutation
                     if (*optarg == '{') {
                         // how many cycles?
@@ -1341,6 +1503,23 @@ invalid_define:
                         goto powerloss_next;
                     }
 
+                    // exhaustive permutations
+                    {
+                        char *parsed = NULL;
+                        size_t count = strtoumax(optarg, &parsed, 0);
+                        if (parsed == optarg) {
+                            goto powerloss_unknown;
+                        }
+                        ((test_powerloss_t*)test_powerlosses)[
+                                test_powerloss_count-1] = (test_powerloss_t){
+                            .run = run_powerloss_exhaustive,
+                            .cycles = NULL,
+                            .cycle_count = count,
+                        };
+                        optarg = (char*)parsed;
+                        goto powerloss_next;
+                    }
+
 powerloss_unknown:
                     // unknown scenario?
                     fprintf(stderr, "error: "