Eric liau
    • Create new note
    • Create a note from template
      • Sharing URL Link copied
      • /edit
      • View mode
        • Edit mode
        • View mode
        • Book mode
        • Slide mode
        Edit mode View mode Book mode Slide mode
      • Customize slides
      • Note Permission
      • Read
        • Only me
        • Signed-in users
        • Everyone
        Only me Signed-in users Everyone
      • Write
        • Only me
        • Signed-in users
        • Everyone
        Only me Signed-in users Everyone
      • Engagement control Commenting, Suggest edit, Emoji Reply
    • Invite by email
      Invitee

      This note has no invitees

    • Publish Note

      Share your work with the world Congratulations! 🎉 Your note is out in the world Publish Note

      Your note will be visible on your profile and discoverable by anyone.
      Your note is now live.
      This note is visible on your profile and discoverable online.
      Everyone on the web can find and read all notes of this public team.
      See published notes
      Unpublish note
      Please check the box to agree to the Community Guidelines.
      View profile
    • Commenting
      Permission
      Disabled Forbidden Owners Signed-in users Everyone
    • Enable
    • Permission
      • Forbidden
      • Owners
      • Signed-in users
      • Everyone
    • Suggest edit
      Permission
      Disabled Forbidden Owners Signed-in users Everyone
    • Enable
    • Permission
      • Forbidden
      • Owners
      • Signed-in users
    • Emoji Reply
    • Enable
    • Versions and GitHub Sync
    • Note settings
    • Note Insights
    • Engagement control
    • Transfer ownership
    • Delete this note
    • Save as template
    • Insert from template
    • Import from
      • Dropbox
      • Google Drive
      • Gist
      • Clipboard
    • Export to
      • Dropbox
      • Google Drive
      • Gist
    • Download
      • Markdown
      • HTML
      • Raw HTML
Menu Note settings Versions and GitHub Sync Note Insights Sharing URL Create Help
Create Create new note Create a note from template
Menu
Options
Engagement control Transfer ownership Delete this note
Import from
Dropbox Google Drive Gist Clipboard
Export to
Dropbox Google Drive Gist
Download
Markdown HTML Raw HTML
Back
Sharing URL Link copied
/edit
View mode
  • Edit mode
  • View mode
  • Book mode
  • Slide mode
Edit mode View mode Book mode Slide mode
Customize slides
Note Permission
Read
Only me
  • Only me
  • Signed-in users
  • Everyone
Only me Signed-in users Everyone
Write
Only me
  • Only me
  • Signed-in users
  • Everyone
Only me Signed-in users Everyone
Engagement control Commenting, Suggest edit, Emoji Reply
  • Invite by email
    Invitee

    This note has no invitees

  • Publish Note

    Share your work with the world Congratulations! 🎉 Your note is out in the world Publish Note

    Your note will be visible on your profile and discoverable by anyone.
    Your note is now live.
    This note is visible on your profile and discoverable online.
    Everyone on the web can find and read all notes of this public team.
    See published notes
    Unpublish note
    Please check the box to agree to the Community Guidelines.
    View profile
    Engagement control
    Commenting
    Permission
    Disabled Forbidden Owners Signed-in users Everyone
    Enable
    Permission
    • Forbidden
    • Owners
    • Signed-in users
    • Everyone
    Suggest edit
    Permission
    Disabled Forbidden Owners Signed-in users Everyone
    Enable
    Permission
    • Forbidden
    • Owners
    • Signed-in users
    Emoji Reply
    Enable
    Import from Dropbox Google Drive Gist Clipboard
       owned this note    owned this note      
    Published Linked with GitHub
    Subscribed
    • Any changes
      Be notified of any changes
    • Mention me
      Be notified of mention me
    • Unsubscribe
    Subscribe
    # 2022q1 Homework1 (lab0) contributed by < [Eric-liau](https://github.com/Eric-liau) > ## 實驗環境 ```shell $ gcc --version gcc (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0 $ lscpu Architecture: x86_64 CPU op-mode(s): 32-bit, 64-bit Byte Order: Little Endian Address sizes: 39 bits physical, 48 bits virtual CPU(s): 8 On-line CPU(s) list: 0-7 Thread(s) per core: 2 Core(s) per socket: 4 Socket(s): 1 NUMA node(s): 1 Vendor ID: GenuineIntel CPU family: 6 Model: 94 Model name: Intel(R) Core(TM) i7-6700HQ CPU @ 2.60GHz Stepping: 3 CPU MHz: 900.002 CPU max MHz: 3500.0000 CPU min MHz: 800.0000 BogoMIPS: 5199.98 Virtualization: VT-x L1d cache: 128 KiB L1i cache: 128 KiB L2 cache: 1 MiB L3 cache: 6 MiB NUMA node0 CPU(s): 0-7 ``` ## [作業要求](https://hackmd.io/@sysprog/linux2022-lab0) 依據指示著手修改 queue.[ch] 和連帶的檔案,測試後用 Git 管理各項修改,記得也該實作 q_sort 函式。 - [x] 在 GitHub 上 fork [lab0-c](https://github.com/sysprog21/lab0-c) - [x] 依據上述指示著手修改 `queue.[ch]` 和連帶的檔案,測試後用 Git 管理各項修改,要能滿足 ``$ make test` 自動評分系統的所有項目。 - [x] 研讀 Linux 核心原始程式碼的 [lib/list_sort.c](https://github.com/torvalds/linux/blob/master/lib/list_sort.c) 並嘗試引入到 `lab0-c` 專案,比較你自己實作的 merge sort 和 Linux 核心程式碼之間效能落差 - [x] 在 `qtest` 提供新的命令 `shuffle`,允許藉由 [Fisher–Yates shuffle](https://en.wikipedia.org/wiki/Fisher%E2%80%93Yates_shuffle) 演算法,對佇列中所有節點進行洗牌 (shuffle) 操作 - [x] 在 `qtest` 提供新的命令 `web`,提供 web 伺服器功能,注意: web 伺服器運作過程中,`qtest` 仍可接受其他命令 - [ ] 開啟 [Address Sanitizer](https://github.com/google/sanitizers/wiki/AddressSanitizer),修正 `qtest` 執行過程中的錯誤 - 先執行 `qtest` 再於命令提示列輸入 `help` 命令,會使開啟 Address Sanitizer 觸發錯誤,應予以排除 - [ ] 運用 Valgrind 排除 `qtest` 實作的記憶體錯誤,並透過 Massif 視覺化 “simulation” 過程中的記憶體使用量,需要設計對應的實驗 - [x] 解釋 [select](https://man7.org/linux/man-pages/man2/select.2.html) 系統呼叫在本程式的使用方式,並分析 [console.c](https://github.com/sysprog21/lab0-c/blob/master/console.c) 的實作,說明其中運用 CS:APP [RIO 套件](http://csapp.cs.cmu.edu/2e/ch10-preview.pdf) 的原理和考量點。可對照參考 [CS:APP 第 10 章重點提示](https://hackmd.io/@sysprog/CSAPP-ch10) - [x] 研讀論文〈[Dude, is my code constant time?](https://eprint.iacr.org/2016/1123.pdf)〉,解釋本程式的 “simulation” 模式是如何透過以實驗而非理論分析,達到驗證時間複雜度,需要解釋 [Student’s t-distribution](https://en.wikipedia.org/wiki/Student%27s_t-distribution) 及程式實作的原理。注意:現有實作存在若干致命缺陷,請討論並提出解決方案; - [ ] 指出現有程式的缺陷 (提示: 和 [RIO 套件](http://csapp.cs.cmu.edu/2e/ch10-preview.pdf) 有關) 或者可改進之處 (例如更全面地使用 Linux 核心風格的鏈結串列 API),嘗試實作並提交 pull request :::danger 針對 2022 年作業要求,更新上述列表 :notes: jserv ::: :::success 已更新 ::: ## queue.c ### q_new - `INIT_LIST_HEAD(head)`: 將 head 的 *\*prev* 和 *\*next* 分別指向 head。 ```c struct list_head *q_new() { struct list_head *head = malloc(sizeof(struct list_head)); if (!head) return NULL; INIT_LIST_HEAD(head); return head; } ``` ### q_free 想法:從 head node 開始,依序把每個`node` free 掉。 - temp : 因為 `container` 會被 free 掉,而 `current` 包含在 `container` 中,所以需要一個 `struct list_head` 來暫存 `current` 的值。 - `free(l)`: `list_for_each` 不會經過 `head` ,所以在迴圈結束後需要把 `head` free 掉。 - 在寫 `q_insert_head` 時提示 memory leak ,檢查後發現` container->value` 並未 free 掉,故加入 `free(container->value)`。 ```c void q_free(struct list_head *l) { struct list_head *current, temp; element_t *container; list_for_each(current, l){ container = list_entry(current, element_t, list); temp = *current; free(container->value); free(container); current = &temp; } free(l); } ``` ### q_insert_head 1. 檢查 `head` 是否為 `null` 2. 分配空間給 `new` 並檢查是否成功分配 3. 分配空間給 `new->value` 並檢查是否成功分配 4. 複製 s 到 `new->value` 5. 將 `new` insert 到 `queue` 中 ```c bool q_insert_head(struct list_head *head, char *s) { if (!head) return false; element_t *new = malloc(sizeof(element_t)); if (!new) return false; new->value = malloc(strlen(s) + 1); if (!new->value) { free(new); return false; } strncpy(new->value, s, strlen(s) + 1); list_add(&new->list, head); return true; } ``` ### q_insert_tail 1. 檢查 `head` 是否為 `null` 2. 分配空間給 `new` 並檢查是否成功分配 3. 分配空間給 `new->value` 並檢查是否成功分配 4. 複製 s 到 `new->value` 5. 將 `new` insert 到 `queue` 中 ```c bool q_insert_tail(struct list_head *head, char *s) { if (!head) return false; element_t *new = malloc(sizeof(element_t)); if (!new) return false; new->value = malloc(strlen(s) + 1); if (!new->value) { free(new); return false; } strncpy(new->value, s, strlen(s) + 1); list_add_tail(&new->list, head); return true; } ``` ### q_remove_head 1. 檢查 `head` 是否為 `NULL` 或 `empty` 2. 檢查 `sp` 是否為 `NULL` 3. 比較 `strlen(node->value) + 1` 和 `bufsize` 大小 - 若後者較大則把 `node->value` 完整複製到 `sp` 中 - 反之則只把 `node->value` 的前 `bufsize - 1` 個字元複製到 `sp`,並在 `sp` 的第 `bufsize` 個字元填上 `'\0' 4. 將 `head->next` 移除 ```c element_t *q_remove_head(struct list_head *head, char *sp, size_t bufsize) { if (!head) return NULL; if (list_empty(head)) return NULL; element_t *node = list_entry(head->next, element_t, list); if (!sp) { list_del_init(head->next); return node; } char *str = node->value; int len = strlen(str) + 1; if (len <= bufsize) strncpy(sp, str, len); else { strncpy(sp, str, bufsize - 1); *(sp + bufsize - 1) = '\0'; } list_del_init(head->next); return node; } ``` ### q_remove_tail 1. 檢查 `head` 是否為 `NULL` 或 `empty` 2. 檢查 `sp` 是否為 `NULL` 3. 比較 `strlen(node->value) + 1` 和 `bufsize` 大小 - 若後者較大則把 `node->value` 完整複製到 `sp` 中 - 反之則只把 `node->value` 的前 `bufsize - 1` 個字元複製到 `sp`,並在 `sp` 的第 `bufsize` 個字元填上 `'\0' 4. 將 `head->prev` 移除 ```c element_t *q_remove_tail(struct list_head *head, char *sp, size_t bufsize) { if (!head) return NULL; if (list_empty(head)) return NULL; element_t *node = list_entry(head->prev, element_t, list); if (!sp) { list_del_init(head->prev); return node; } char *str = node->value; int len = strlen(str) + 1; if (len <= bufsize) strncpy(sp, str, len); else { strncpy(sp, str, bufsize - 1); *(sp + bufsize - 1) = '\0'; } list_del_init(head->prev); return node; } ``` ### q_release_element 未做更動 ```c void q_release_element(element_t *e) { free(e->value); free(e); } ``` ### q_size 整個 list 走過一遍即可算出 size ```c int q_size(struct list_head *head) { if (!head) return 0; int len = 0; struct list_head *li; list_for_each (li, head) len++; return len; } ``` ### q_delete_mid 利用快慢指標找到中間的 node ```c struct list_head *find_mid(struct list_head *head) { struct list_head *slow = head->next; for (struct list_head *fast = head->next; fast != head && fast->next != head; fast = fast->next->next) { slow = slow->next; } return slow; } bool q_delete_mid(struct list_head *head) { if (!head) return false; if (list_empty(head)) return false; struct list_head *mid = find_mid(head); mid->prev->next = mid->next; mid->next->prev = mid->prev; element_t *node = list_entry(mid, element_t, list); free(node->value); free(node); return true; } ``` ### q_delete_dup 若 `current node` 的值等於 `first` 的值就把 `last` 指標移到 `current node` ,若不等於就檢查 `first` 和 `last` 是否為同一個 `node` ,若不是就把 `first` 和 `last` 之間的 `node` 全部刪除。 ```c bool q_delete_dup(struct list_head *head) { if (!head) return false; if (list_empty(head)) return true; element_t *current, *first = list_entry(head->next, element_t, list), *last = first; list_for_each_entry (current, head, list) { if (!strcmp(current->value, first->value)) last = current; else { if (first != last) { first->list.prev->next = last->list.next; last->list.next->prev = first->list.prev; while (first != current) { element_t *temp = list_entry(first->list.next, element_t, list); q_release_element(first); first = temp; } } first = current; last = current; } } if (first != last) { first->list.prev->next = last->list.next; last->list.next->prev = first->list.prev; while (first != current) { element_t *temp = list_entry(first->list.next, element_t, list); q_release_element(first); first = temp; } } return true; } ``` ### q_swap 把 list 的 node 每兩個依序前後交換 ```c void q_swap(struct list_head *head) { if (!head) return; for (struct list_head *first = head->next, *second = head->next->next; first != head && second != head; first = first->next, second = first->next) { struct list_head *prev = first->prev; struct list_head *next = second->next; prev->next = second; second->prev = prev; second->next = first; first->prev = second; next->prev = first; first->next = next; } } ``` ### q_reverse 把每個 node 的 `next` 和 `prev` 進行交換 ```c void q_reverse(struct list_head *head) { if (!head) return; struct list_head *current, *safe; list_for_each_safe (current, safe, head) { struct list_head *prev = current->prev; struct list_head *next = current->next; current->next = prev; current->prev = next; } struct list_head *prev = head->prev; struct list_head *next = head->next; head->next = prev; head->prev = next; } ``` ### q_sort #### 遞迴版本 使用遞迴的 merge sort ,因為傳入的 list 有一個不存資料的 `head` ,所以宣告一個 `new_head` 作為切下來另一半 list 的 `head node`。在自己電腦的效能測試過了,但 github 上的仍無法通過。 ```c struct list_head *merge(struct list_head *left, struct list_head *right) { struct list_head head, **temp, *l = left->next, *r = right->next; INIT_LIST_HEAD(&head); while (l != left && r != right) { if (strcmp(list_entry(l, element_t, list)->value, list_entry(r, element_t, list)->value) < 0) temp = &l; else temp = &r; *temp = (*temp)->next; list_move_tail((*temp)->prev, &head); } (l == left) ? list_splice_tail_init(right, &head) : list_splice_tail(left, &head); list_splice_tail(&head, right); return right; } void q_sort(struct list_head *head) { if (!head || list_empty(head) || list_is_singular(head)) return; struct list_head *mid = find_mid(head); struct list_head new_head; INIT_LIST_HEAD(&new_head); list_cut_position(&new_head, head, mid->prev); q_sort(&new_head); q_sort(head); merge(&new_head, head); return; } ``` - 參考[這篇筆記](https://hackmd.io/@laneser/linux2022_lab0#q_sort)懷疑是 stackoverflow 導致效能測試無法通過 - 思考為什麼會發生 stackoverflow ,猜測是因為每次遞迴都會宣告一次 `new_head` 所導致的 - 改寫 mergesort ,讓 merge 的時候不考慮 `head node` ,如此就不需要每次遞迴都宣告一個 `new_head` 。 - 再次上傳後成功通過效能測試 ```c struct list_head *merge(struct list_head *left, struct list_head *right) { struct list_head *head = NULL, **temp = &head; while (left && right) { if (strcmp(list_entry(left, element_t, list)->value, list_entry(right, element_t, list)->value) <= 0) { *temp = left; left = left->next; } else { *temp = right; right = right->next; } temp = &(*temp)->next; } *temp = left ? left : right; return head; } struct list_head *merge_sort(struct list_head *head) { if (!head) return head; if (!head->next) return head; struct list_head *fast = head, *slow = head; while (1) { if (!fast) break; if (!fast->next) break; slow = slow->next; fast = fast->next->next; } slow->prev->next = NULL; head = merge_sort(head); slow = merge_sort(slow); return merge(head, slow); } void q_sort(struct list_head *head) { if (!head || list_empty(head) || list_is_singular(head)) return; head->prev->next = NULL; head->next = merge_sort(head->next); struct list_head *current, *prev; for (current = head->next, prev = head; current; current = current->next, prev = prev->next) current->prev = prev; prev->next = head; head->prev = prev; return; } ``` #### 非遞迴版本 依序將 list 裡的 node 存進 `space` 陣列中,當第 `space[i]` 與 `space[i-1]` 擁有同樣的 node 數(lsit 的長度一樣)時,將兩者進行 merge 並存到 `space[i-1]` 。 ```c void merge_sort_iter(struct list_head *head) { int num = q_size(head); num = log2(num) + 1; int stack[num]; struct list_head *space[num]; head->prev->next = NULL; head->prev = NULL; struct list_head *node = head->next; int i = 0; while (node) { space[i] = node; stack[i] = 0; struct list_head *tmp = node; node = node->next; tmp->next = NULL; int j = i - 1; while (j >= 0) { if (stack[j] == stack[j + 1]) { space[j] = merge(space[j], space[j + 1]); stack[j]++; j--; i--; } else break; } i++; } i--; while (i) { space[i - 1] = merge(space[i - 1], space[i]); i--; } head->next = space[0]; struct list_head *current, *prev; for (current = head->next, prev = head; current; current = current->next, prev = prev->next) current->prev = prev; prev->next = head; head->prev = prev; return; } ``` ### 研讀 [linux/list_sort.c](https://github.com/torvalds/linux/blob/master/lib/list_sort.c) - 需要傳入 `list_sort()` 的三個參數 1. `priv` : 傳輸 `cmp` 所需參數的 pointer 2. `head` : 要排序的 list 3. `cmp` : 比較自定義大小的 function pointer,當 `a` 須置於 `b` 之後時會回傳大於 0 的值 - 以升序排列為例,當 `a` > `b` 時便會回傳大於 0 的值 - `list_sort` 為 stable sort ,故不必區分小於或等於 0 的分別 - `list_sort` 跟一般的 `bottom-up mergesort` 不同,後者每當出現 2 個 $2^k$ 大小的 list 時就會進行合併,前者則是當第 3 個 $2^k$ 大小的 list 出現時才進行合併。這樣做的好處是避免了 worst case 的出現,當要排序的 list 大小只比 $2^k$ 大一點點時,`bottom-up mergesort` 的最後一次合併就會出現極端的情況。舉個例子,當 list 大小為 9 時, `bottom-up mergesort`最後一次合併的 2 個 list 大小分別會是 8 和 1;而 `list_sort` 則會是 4 和 5 ,後者很明顯地會優於前者。事實上,不論要排序的 list 大小為何, `list_sort` 的合併都不會出現比 2 : 1 還糟糕的情況。 - 從 `list_sort` 的程式碼中,我發現了幾個自己實作的 merge sort 可以改進的地方 - sort 過的 list 可以用 `struct list_head()` 的 `prev` 指標來串聯,如此就不須宣告新的空間去存放 - 試著只用一個 integer `count` 去控制 list 的合併 - 上述兩點可以使 sort 少走訪一次整個 list (不必在一開始計算 list 的 node 數) - 在最後一次的合併就把整個 list 的 `prev` 連接回去 - 可以使 sort 少走訪一次 list - 修改後的程式碼 ```c void final_merge(struct list_head *left, struct list_head *right, struct list_head *head) { struct list_head **temp = &head; while (left && right) { if (strcmp(list_entry(left, element_t, list)->value, list_entry(right, element_t, list)->value) <= 0) { left->prev = *temp; (*temp)->next = left; left = left->next; } else { right->prev = *temp; (*temp)->next = right; right = right->next; } temp = &(*temp)->next; } (*temp)->next = left ? left : right; while ((*temp)->next) { (*temp)->next->prev = *temp; temp = &(*temp)->next; } (*temp)->next = head; head->prev = *temp; return; } void merge_sort_iter(struct list_head *head) { struct list_head *pending = NULL; struct list_head *list = head->next; head->prev->next = NULL; head->prev = NULL; int count = 0; while (list) { list->prev = pending; pending = list; list = list->next; pending->next = NULL; struct list_head **tail = &pending; for (int bits = count;; bits >>= 1) { if (bits & 1) { struct list_head *a = *tail, *b = a->prev; a = merge(b, a); a->prev = b->prev; *tail = a; } else break; } count++; } list = pending; pending = pending->prev; while (pending) { struct list_head *next = pending->prev; if (!next) break; list = merge(pending, list); pending = next; } final_merge(pending, list, head); return; } ``` - 測試兩者間效能差異 - 測試用的命令檔(由於第一次 sort 測出來的時間會跟之後的有明顯落差,故皆不採記第一次的結果) ``` option echo 0 option verbose 1 # fault new it RAND 300000 time sort # 1 new it RAND 300000 time sort # 2 new it RAND 300000 time sort # 3 new it RAND 300000 time sort # 4 new it RAND 300000 time sort # 5 new it RAND 300000 time sort ``` - 測試結果 - 原始版本 ``` ericliau@ericliau-GL502VML:~/linux2022/lab0-c$ ./qtest -f traces/test1.cmd cmd> option echo 0 # fault Delta time = 0.242 # 1 Delta time = 0.381 # 2 Delta time = 0.396 # 3 Delta time = 0.391 # 4 Delta time = 0.386 # 5 Delta time = 0.383 ``` - 修改後 ``` ericliau@ericliau-GL502VML:~/linux2022/lab0-c$ ./qtest -f traces/test1.cmd cmd> option echo 0 # fault Delta time = 0.207 # 1 Delta time = 0.325 # 2 Delta time = 0.327 # 3 Delta time = 0.333 # 4 Delta time = 0.333 # 5 Delta time = 0.330 ``` - 把修改後的 5 次測試時間加起來除以原始版本的 5 次測試時間,得出來的值約為 0.85 ,效能提昇了 15% 左右 ### 引入 [linux/list_sort.c](https://github.com/torvalds/linux/blob/master/lib/list_sort.c) 引入方法參考自 [2020leon](https://hackmd.io/@6649/linux2022-lab0#%E5%AF%A6%E5%81%9A%E4%BD%87%E5%88%97%EF%BC%88-queuec-%EF%BC%89) 1. 將 `list_sort.c` 和 `list_sort.h` 複製到 lab0-c 專案中。 - `list_sort.c` 直接從 [linux/list_sort.c](https://github.com/torvalds/linux/blob/master/lib/list_sort.c) 複製。 - `list_sort.h` 從 linux 核心中的 include/linux/list_sort.h 複製。 2. 將型別 `u8` 更改為 `uint8_t`。 - 必須引入標頭檔 `stdint.h`。 3. 在 `list_sort.h` 加入 `likely(x)` 及 `unlikely(x)` 巨集定義。 ```c # define likely(x) __builtin_expect(!!(x), 1) # define unlikely(x) __builtin_expect(!!(x), 0) ``` 4. 在 `list_sort.c` 中新增 `listcmp()` 函式。 ```c int listcmp(void *priv, const struct list_head *a, const struct list_head *b) { return strcmp(list_entry(a, element_t, list)->value, list_entry(b, element_t, list)->value); } ``` 5. 使 cppcheck 忽略 `list_sort.c` 中 `merge` 函式產生的 `head` 未賦值錯誤 - 加入 `// cppcheck-suppress unassignedVariable` ```c // cppcheck-suppress unassignedVariable struct list_head *head, **tail = &head; ``` - 於終端機輸入 `$ cppcheck --inline-suppr list_sort.c` 6. 修改 `qtest.c` ,新增一個 option 以便於在 `list_sort` 及自己實作的排序法間做切換。 - 加入 option ```c ... static int listsort = 0; ... static void console_init() { ... add_param("listsort", &listsort, "Use list_sort or not", NULL); } ``` - 修改 `do_sort()` 函式 ```c bool do_sort(int argc, char *argv[]) { ... if (exception_setup(true)){ if(!listsort) q_sort(l_meta.l); else list_sort(NULL, l_meta.l, cmp); ... } ``` 測試 list sort 和自己的排序法效能落差,測出來自己的排序法慢了大概 10% ``` ericliau@ericliau-GL502VML:~/linux2022/lab0-c$ ./qtest -f traces/test.cmd cmd> option echo 0 # fault Delta time = 0.273 # user version Delta time = 0.376 Delta time = 0.408 Delta time = 0.390 Delta time = 0.377 Delta time = 0.374 # list sort Delta time = 0.358 Delta time = 0.342 Delta time = 0.340 Delta time = 0.342 Delta time = 0.344 ``` 而如果是用參考 list sort 後所修改的版本則大約快了 5% ``` ericliau@ericliau-GL502VML:~/linux2022/lab0-c$ ./qtest -f traces/test.cmd cmd> option echo 0 # fault Delta time = 0.207 # user version Delta time = 0.323 Delta time = 0.321 Delta time = 0.327 Delta time = 0.336 Delta time = 0.333 # list sort Delta time = 0.344 Delta time = 0.348 Delta time = 0.344 Delta time = 0.343 Delta time = 0.343 ``` 我還蠻意外會比 list sort 快的,具體比較快的原因待驗證 ## qtest.c ### 實作 shuffle 演算法 使用 [Fisher–Yates shuffle](https://en.wikipedia.org/wiki/Fisher%E2%80%93Yates_shuffle) 演算法,該演算法之 pseudocode 如下: ``` -- To shuffle an array a of n elements (indices 0..n-1): for i from n−1 downto 1 do j ← random integer such that 0 ≤ j ≤ i exchange a[j] and a[i] ``` 實作之程式碼如下: ```c static void shuffle(struct list_head *head) { int num = q_size(head); struct list_head *tmp, *current; for(tmp = head->prev; num > 0; num--){ current = head; for(int i = rand() % num; i >= 0; i--) current = current->next; char *value = list_entry(current, element_t, list)->value; list_entry(current, element_t, list)->value = list_entry(tmp, element_t, list)->value; list_entry(tmp, element_t, list)->value = value; } } static bool do_shuffle(int argc, char *argv[]) { if (argc != 1) { report(1, "%s takes no arguments", argv[0]); return false; } if(!l_meta.l){ report(1, "list is null"); return false; } else if (list_empty(l_meta.l)) { report(1, "list is empty"); return false; } else if (list_is_singular(l_meta.l)) { report(1, "no need to shuffle"); return false; } shuffle(l_meta.l); show_queue(3); return true; } ``` 最後在 `console_init()` 添加指令: ``` ADD_COMMAND(shuffle, " | Shuffle the list"); ``` ### 提供 web 命令 因為要整合 [tiny-web-server](https://github.com/7890/tiny-web-server) ,所以先試著引入 `tiny.c` ,並分一個 `tiny.h` 出來存放其他檔案會用到的部份。 - 其中的 `process()` 需要進行修改,這部份稍後再提 ```c #ifndef __TINY_H #define __TINY_H #include <arpa/inet.h> /* inet_ntoa */ #include <dirent.h> #include <errno.h> #include <fcntl.h> #include <netinet/in.h> #include <netinet/tcp.h> #include <signal.h> #include <stdio.h> #include <stdlib.h> #include <string.h> #include <sys/sendfile.h> #include <sys/socket.h> #include <sys/stat.h> #include <sys/types.h> #include <time.h> #include <unistd.h> typedef struct { char filename[512]; off_t offset; /* for support Range */ size_t end; } http_request; /* Simplifies calls to bind(), connect(), and accept() */ typedef struct sockaddr SA; int open_listenfd(int port); char *process(int fd, struct sockaddr_in *clientaddr); void parse_request(int fd, http_request *req); #endif /* __TINY_H */ ``` 然後把 `tiny.c` 和 `tiny.h` 加到 `Makefile` 裡,在 `OBJS := `後面加入 `tiny.o` ``` OBJS := console.o qtest.o report.o harness.o queue.o \ random.o dudect/constant.o dudect/fixture.o dudect/ttest.o \ linenoise.o tiny.o ``` 接著基本上就照著[作業說明](https://hackmd.io/@sysprog/linux2022-lab0#%E6%95%B4%E5%90%88-tiny-web-server)完成 首先在 `linenoise.h` 中加入變數 `noise` 和 `listenfd` ```c bool noise; int listenfd; ``` 修改 `linenoiseEdit()` ,在 `while(1)` 裡加入以下程式碼 ```c if (listenfd) { fd_set set; FD_ZERO(&set); FD_SET(listenfd, &set); FD_SET(stdin_fd, &set); struct sockaddr_in clientaddr; socklen_t clientlen = sizeof(clientaddr); int rv = select(listenfd + 1, &set, NULL, NULL, NULL); int connfd; switch (rv) { case -1: perror("select"); /* an error occurred */ continue; case 0: printf("timeout occurred\n"); /* a timeout occurred */ continue; default: if (FD_ISSET(listenfd, &set)) { connfd = accept(listenfd, (SA *) &clientaddr, &clientlen); char *p = process(connfd, &clientaddr); strncpy(buf, p, strlen(p) + 1); close(connfd); free(p); return strlen(p); } else if (FD_ISSET(stdin_fd, &set)) { nread = read(l.ifd, &c, 1); if (nread <= 0) return l.len; } break; } } else { nread = read(l.ifd, &c, 1); if (nread <= 0) return l.len; } ``` 修改 `run_console()` ,讓其依照 `linenoise` 開啟與否決定要用 `linenoise` 還是 `cmd_select` ```c bool run_console(char *infile_name) { if (!push_file(infile_name)) { report(1, "ERROR: Could not open source file '%s'", infile_name); return false; } if (!has_infile) { char *cmdline; while (noise && (cmdline = linenoise(prompt)) != NULL) { interpret_cmd(cmdline); linenoiseHistoryAdd(cmdline); /* Add to the history. */ linenoiseHistorySave(HISTORY_FILE); /* Save the history on disk. */ linenoiseFree(cmdline); } if (!noise) { while (!cmd_done()) { cmd_select(0, NULL, NULL, NULL, NULL); } } } else { while (!cmd_done()) { cmd_select(0, NULL, NULL, NULL, NULL); } } return err_cnt == 0; } ``` 修改 `cmd_select()` ,新增對於 web 端輸入的處理 ```c int cmd_select(int nfds, fd_set *readfds, fd_set *writefds, fd_set *exceptfds, struct timeval *timeout) { int infd; fd_set local_readset; if (cmd_done()) return 0; if (!block_flag) { /* Process any commands in input buffer */ if (!readfds) readfds = &local_readset; /* Add input fd to readset for select */ infd = buf_stack->fd; FD_ZERO(readfds); FD_SET(infd, readfds); /* If web not ready listen */ if (listenfd != -1) FD_SET(listenfd, readfds); if (infd == STDIN_FILENO && prompt_flag) { printf("%s", prompt); fflush(stdout); prompt_flag = true; } if (infd >= nfds) nfds = infd + 1; if (listenfd >= nfds) nfds = listenfd + 1; } if (nfds == 0) return 0; int result = select(nfds, readfds, writefds, exceptfds, timeout); if (result <= 0) return result; infd = buf_stack->fd; if (readfds && FD_ISSET(infd, readfds)) { /* Commandline input available */ FD_CLR(infd, readfds); result--; char *cmdline; cmdline = readline(); if (cmdline) interpret_cmd(cmdline); } else if (readfds && FD_ISSET(listenfd, readfds)) { FD_CLR(listenfd, readfds); result--; int connfd; struct sockaddr_in clientaddr; socklen_t clientlen = sizeof(clientaddr); connfd = accept(listenfd, (SA *) &clientaddr, &clientlen); char *p = process(connfd, &clientaddr); if (p) interpret_cmd(p); free(p); close(connfd); } return result; } ``` 修改 `tiny.c` 中的 `process()` , `process()` 會處理 http 請求,將其轉換成符合 `cmd` 的命令格式 ```c char *process(int fd, struct sockaddr_in *clientaddr) { #ifdef LOG_ACCESS printf("accept request, fd is %d, pid is %d\n", fd, getpid()); #endif http_request req; parse_request(fd, &req); int status = 200; char *p = req.filename; /* Change '/' to ' ' */ while (*p && (*p) != '\0') { ++p; if (*p == '/') { *p = ' '; } } #ifdef LOG_ACCESS log_access(status, clientaddr, &req); #endif char *ret = malloc(strlen(req.filename) + 1); strncpy(ret, req.filename, strlen(req.filename) + 1); return ret; } ``` 最後在 `qtest.c` 中新增 `do_web()` 函式和 `web` 指令 - 加入 `set_echo(false);` 是為了避免在 `web` 開啟後從使用者端輸入指令會重複印出的情形 ```c static bool do_web(int argc, char *argv[]) { noise = false; set_echo(false); listenfd = open_listenfd(9999); printf("listen on port 9999, fd is %d\n", listenfd); return true; } ``` ```c ADD_COMMAND(web, " | Open web server"); ``` 執行結果 ``` ericliau@ericliau-GL502VML:~/linux2022/lab0-c$ ./qtest cmd> web listen on port 9999, fd is 3 cmd> accept request, fd is 4, pid is 53178 127.0.0.1:58584 200 - 'new' (text/plain) l = [] cmd> accept request, fd is 4, pid is 53178 127.0.0.1:58586 200 - 'ih a' (text/plain) l = [a] cmd> accept request, fd is 4, pid is 53178 127.0.0.1:58588 200 - 'ih b' (text/plain) l = [b a] cmd> accept request, fd is 4, pid is 53178 127.0.0.1:58590 200 - 'ih c' (text/plain) l = [c b a] cmd> ih d l = [d c b a] cmd> ih e l = [e d c b a] cmd> accept request, fd is 4, pid is 53178 127.0.0.1:58592 200 - 'sort' (text/plain) l = [a b c d e] ``` ## 開啟 [Address Sanitizer](https://github.com/google/sanitizers/wiki/AddressSanitizer),修正 `qtest` 執行過程中的錯誤 `qtest` 執行後並未發生錯誤,尚在尋找原因 ## select 系統呼叫 `select` 的功用是同時監測多個 `file descriptor` ,直到其中一個準備好要進行 I/O 。而 `lab0` 使用 `select` 的地方在函式 `cmd_select()` 中,透過實作 `web` 指令,可以得知其主要用來偵測是由使用者端還是從 web 端輸入指令。 ## 分析 console.c 先說結論 : 使用 RIO 套件可以在讀 cmd 檔的時候減少呼叫系統指令 `read` 的次數。 因為 RIO 套件支援 buffered I/O ,在讀行時只需要呼叫 1 次 `read` 系統指令將一個 buffer 大小的資料存進 buffer ,再從 buffer 中一個字一個字的尋找換行符即可。如果不使用 buffer ,因為需要尋找換行符的緣故,每次呼叫 `read` 只能讀一個字元,導致需要頻繁的呼叫 `read` 指令,使得 OS 必須不斷地在 kernel-mode 和 user-mode 間做切換。 ## 研讀論文〈Dude, is my code constant time?〉 這篇論文使用 [dudect](https://github.com/oreparaz/dudect) 這個程式來測量某個程式執行時間是否為 constant time 。其中使用的 [Student’s t-distribution](https://en.wikipedia.org/wiki/Student%27s_t-distribution) 為一種估計期望值的統計學方法,用於樣本量不大且不知道標準差的情況。 該論文使用 leakage detection test ,因為是透過統計模型來推斷是否為 constant time,使得此方法並不局限於特定的 CPU 硬體,方法步驟如下: 1. 分別用固定和隨機的輸入資料進行多次測試 2. 去掉極端值然後進行 high-order preprocessing 3. 用統計方法判定是否符合常態時間,本專案使用 [Welch's t-test](https://en.wikipedia.org/wiki/Welch's_t-test) ### lab0 實作 lab0 檢查是否 constant time 的函式主要位於 `dudect/` 中,其呼叫路徑為 `is_XXX_const() -> TEST_CONST() -> doit()` ,測試步驟基本在 `doit()` 中完成。 ```c static bool doit(int mode) { int64_t *before_ticks = calloc(n_measure + 1, sizeof(int64_t)); int64_t *after_ticks = calloc(n_measure + 1, sizeof(int64_t)); int64_t *exec_times = calloc(n_measure, sizeof(int64_t)); uint8_t *classes = calloc(n_measure, sizeof(uint8_t)); uint8_t *input_data = calloc(n_measure * chunk_size, sizeof(uint8_t)); if (!before_ticks || !after_ticks || !exec_times || !classes || !input_data) { die(); } prepare_inputs(input_data, classes); measure(before_ticks, after_ticks, input_data, mode); differentiate(exec_times, before_ticks, after_ticks); update_statistics(exec_times, classes); bool ret = report(); free(before_ticks); free(after_ticks); free(exec_times); free(classes); free(input_data); return ret; } ``` `prepare_inputs()` 產生多個長度為 7 bytes 的隨機字串 (加上 `\0` 為 8 bytes),並將資料分成兩個 class。 `measure()` 將 `prepare_inputs()` 產生的字串根據不同操作進行測量,測量方式為分別記錄執行該操作前後當下的 cycle 數。 `differentiate()` 計算兩者 cycle 數的差值,即該操作執行所花費的 cycle 數。 `update_statistics()` 再用得到的執行 cycle 數分別算出兩個 class 的平均數和變異數。 `report()` 將所求出的值用 Welch’s t-test 算出 `t` 值,並判斷對應操作是否為 constant time 。

    Import from clipboard

    Paste your markdown or webpage here...

    Advanced permission required

    Your current role can only read. Ask the system administrator to acquire write and comment permission.

    This team is disabled

    Sorry, this team is disabled. You can't edit this note.

    This note is locked

    Sorry, only owner can edit this note.

    Reach the limit

    Sorry, you've reached the max length this note can be.
    Please reduce the content or divide it to more notes, thank you!

    Import from Gist

    Import from Snippet

    or

    Export to Snippet

    Are you sure?

    Do you really want to delete this note?
    All users will lose their connection.

    Create a note from template

    Create a note from template

    Oops...
    This template has been removed or transferred.
    Upgrade
    All
    • All
    • Team
    No template.

    Create a template

    Upgrade

    Delete template

    Do you really want to delete this template?
    Turn this template into a regular note and keep its content, versions, and comments.

    This page need refresh

    You have an incompatible client version.
    Refresh to update.
    New version available!
    See releases notes here
    Refresh to enjoy new features.
    Your user state has changed.
    Refresh to load new user state.

    Sign in

    Forgot password

    or

    By clicking below, you agree to our terms of service.

    Sign in via Facebook Sign in via Twitter Sign in via GitHub Sign in via Dropbox Sign in with Wallet
    Wallet ( )
    Connect another wallet

    New to HackMD? Sign up

    Help

    • English
    • 中文
    • Français
    • Deutsch
    • 日本語
    • Español
    • Català
    • Ελληνικά
    • Português
    • italiano
    • Türkçe
    • Русский
    • Nederlands
    • hrvatski jezik
    • język polski
    • Українська
    • हिन्दी
    • svenska
    • Esperanto
    • dansk

    Documents

    Help & Tutorial

    How to use Book mode

    Slide Example

    API Docs

    Edit in VSCode

    Install browser extension

    Contacts

    Feedback

    Discord

    Send us email

    Resources

    Releases

    Pricing

    Blog

    Policy

    Terms

    Privacy

    Cheatsheet

    Syntax Example Reference
    # Header Header 基本排版
    - Unordered List
    • Unordered List
    1. Ordered List
    1. Ordered List
    - [ ] Todo List
    • Todo List
    > Blockquote
    Blockquote
    **Bold font** Bold font
    *Italics font* Italics font
    ~~Strikethrough~~ Strikethrough
    19^th^ 19th
    H~2~O H2O
    ++Inserted text++ Inserted text
    ==Marked text== Marked text
    [link text](https:// "title") Link
    ![image alt](https:// "title") Image
    `Code` Code 在筆記中貼入程式碼
    ```javascript
    var i = 0;
    ```
    var i = 0;
    :smile: :smile: Emoji list
    {%youtube youtube_id %} Externals
    $L^aT_eX$ LaTeX
    :::info
    This is a alert area.
    :::

    This is a alert area.

    Versions and GitHub Sync
    Get Full History Access

    • Edit version name
    • Delete

    revision author avatar     named on  

    More Less

    Note content is identical to the latest version.
    Compare
      Choose a version
      No search result
      Version not found
    Sign in to link this note to GitHub
    Learn more
    This note is not linked with GitHub
     

    Feedback

    Submission failed, please try again

    Thanks for your support.

    On a scale of 0-10, how likely is it that you would recommend HackMD to your friends, family or business associates?

    Please give us some advice and help us improve HackMD.

     

    Thanks for your feedback

    Remove version name

    Do you want to remove this version name and description?

    Transfer ownership

    Transfer to
      Warning: is a public team. If you transfer note to this team, everyone on the web can find and read this note.

        Link with GitHub

        Please authorize HackMD on GitHub
        • Please sign in to GitHub and install the HackMD app on your GitHub repo.
        • HackMD links with GitHub through a GitHub App. You can choose which repo to install our App.
        Learn more  Sign in to GitHub

        Push the note to GitHub Push to GitHub Pull a file from GitHub

          Authorize again
         

        Choose which file to push to

        Select repo
        Refresh Authorize more repos
        Select branch
        Select file
        Select branch
        Choose version(s) to push
        • Save a new version and push
        • Choose from existing versions
        Include title and tags
        Available push count

        Pull from GitHub

         
        File from GitHub
        File from HackMD

        GitHub Link Settings

        File linked

        Linked by
        File path
        Last synced branch
        Available push count

        Danger Zone

        Unlink
        You will no longer receive notification when GitHub file changes after unlink.

        Syncing

        Push failed

        Push successfully