From d1cd99244f96bd95bbd67cc737248df346dec08b Mon Sep 17 00:00:00 2001 From: egor-tensin Date: Thu, 25 Apr 2024 03:50:51 +0000 Subject: =?UTF-8?q?Deploying=20to=20gh-pages=20from=20@=20egor-tensin/cimp?= =?UTF-8?q?le@8e652dd2cb69928ea1596aa3e59845fef6854e2c=20=F0=9F=9A=80?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- ..._server.c.68fbb5aebb39dbc263f5110d228c9b35.html | 2532 ++++++++++++++++++++ 1 file changed, 2532 insertions(+) create mode 100644 coverage/index.tcp_server.c.68fbb5aebb39dbc263f5110d228c9b35.html (limited to 'coverage/index.tcp_server.c.68fbb5aebb39dbc263f5110d228c9b35.html') diff --git a/coverage/index.tcp_server.c.68fbb5aebb39dbc263f5110d228c9b35.html b/coverage/index.tcp_server.c.68fbb5aebb39dbc263f5110d228c9b35.html new file mode 100644 index 0000000..d3df3c6 --- /dev/null +++ b/coverage/index.tcp_server.c.68fbb5aebb39dbc263f5110d228c9b35.html @@ -0,0 +1,2532 @@ + + + + + + GCC Code Coverage Report + + + + + +

GCC Code Coverage Report

+ +
+ +
+
+ + + + + + + + + + + + + +
Directory:src/
File:src/tcp_server.c
Date:2024-04-25 03:45:42
+
+
+ + + + + + + + + + + + + + + + + + + +
ExecTotalCoverage
Lines:8712271.3%
Branches:276640.9%
+
+
+ +
+ +
+ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
LineBranchExecSource
1 + /*
2 + * Copyright (c) 2022 Egor Tensin <egor@tensin.name>
3 + * This file is part of the "cimple" project.
4 + * For details, see https://github.com/egor-tensin/cimple.
5 + * Distributed under the MIT License.
6 + */
7 +
8 + #include "tcp_server.h"
9 + #include "compiler.h"
10 + #include "event_loop.h"
11 + #include "file.h"
12 + #include "log.h"
13 + #include "net.h"
14 + #include "signal.h"
15 +
16 + #include <poll.h>
17 + #include <pthread.h>
18 + #include <signal.h>
19 + #include <stdlib.h>
20 + #include <sys/eventfd.h>
21 + #include <sys/queue.h>
22 + #include <unistd.h>
23 +
24 + /*
25 + * This is a simple threaded TCP server implementation. Each client is handled
26 + * in a separate thread.
27 + *
28 + * It used to be much simpler; basically, we have two types of client
29 + * connections: those made by cimple-worker and cimple-client respectively.
30 + * cimple-server would keep track of cimple-worker threads/connections, and
31 + * clean them up when assigning tasks/on shutdown.
32 + *
33 + * What about cimple-client connections though? I struggled to come up with a
34 + * scheme that would allow cimple-server to clean them up gracefully. When
35 + * would it do the cleanup even? I didn't want to do it on shutdown, since
36 + * there would be potentially a lot of them.
37 + *
38 + * One solution is to make client threads detached. This is a common advise;
39 + * I really don't understand the merit of this approach though. Client threads
40 + * actively work on shared data, take locks, etc. Data corruption is very
41 + * likely after the main thread exits and all the rest are killed.
42 + *
43 + * Another approach is pre-threading; we make a number of threads beforehand
44 + * and handle all client connections; I view this approach as limiting in
45 + * principle; probably that's foolish of me.
46 + *
47 + * Finally, I cannot bring myself to do non-blocking I/O. I honestly fear the
48 + * amount of work it would require to maintain read buffers, etc.
49 + *
50 + * So I came up with this convoluted scheme. The TCP server adds the listening
51 + * socket to the event loop, as before. Each client thread makes an eventfd
52 + * descriptor that it writes to when it's about to finish. The eventfd
53 + * descriptor is added to the event loop; once it's readable, we clean up the
54 + * client thread quickly from the main event loop thread. The TCP server itself
55 + * keeps track of client threads; on shutdown, it cleans up those still working.
56 + *
57 + * I'm _really_ not sure about this approach, it seems fishy as hell; I guess,
58 + * we'll see.
59 + */
60 +
61 + struct client {
62 + struct tcp_server *server;
63 + int conn_fd;
64 +
65 + int cleanup_fd;
66 +
67 + pid_t tid;
68 + pthread_t thread;
69 +
70 + SIMPLEQ_ENTRY(client) entries;
71 + };
72 +
73 + SIMPLEQ_HEAD(client_queue, client);
74 +
75 + struct tcp_server {
76 + struct event_loop *loop;
77 +
78 + tcp_server_conn_handler conn_handler;
79 + void *conn_handler_arg;
80 +
81 + struct client_queue client_queue;
82 +
83 + int accept_fd;
84 + };
85 +
86 + 27620static void client_destroy(struct client *client)
87 + {
88 +
+ 1/2 +
+
✓ Branch 1 taken 27620 times.
+
✗ Branch 2 not taken.
+
+
+ 		27620 log_debug("Cleaning up client thread %d\n", client->tid);
89 +
90 +
+ 8/8 +
+
✓ Branch 0 taken 14983 times.
+
✓ Branch 1 taken 12637 times.
+
✓ Branch 2 taken 6743 times.
+
✓ Branch 3 taken 8240 times.
+
✓ Branch 4 taken 8078 times.
+
✓ Branch 5 taken 12637 times.
+
✓ Branch 6 taken 9276 times.
+
✓ Branch 7 taken 3361 times.
+
+
+ 		35698 SIMPLEQ_REMOVE(&client->server->client_queue, client, client, entries);
91 +
+ 1/4 +
+
✗ Branch 1 not taken.
+
✓ Branch 2 taken 27620 times.
+
✗ Branch 4 not taken.
+
✗ Branch 5 not taken.
+
+
+ 		27620 pthread_errno_if(pthread_join(client->thread, NULL), "pthread_join");
92 + 27620 file_close(client->cleanup_fd);
93 + 27620 net_close(client->conn_fd);
94 + 27620 free(client);
95 + 27620}
96 +
97 + 27620static int client_destroy_handler(UNUSED struct event_loop *loop, UNUSED int fd,
98 + UNUSED short revents, void *_client)
99 + {
100 + 27620 struct client *client = (struct client *)_client;
101 +
+ 1/2 +
+
✓ Branch 1 taken 27620 times.
+
✗ Branch 2 not taken.
+
+
+ 		27620 log_debug("Client thread %d indicated that it's done\n", client->tid);
102 +
103 + 27620 client_destroy(client);
104 + 27620 return 0;
105 + }
106 +
107 + 27620static void *client_thread_func(void *_client)
108 + {
109 + 27620 struct client *client = (struct client *)_client;
110 + 27620 int ret = 0;
111 +
112 + 27620 client->tid = gettid();
113 +
+ 1/2 +
+
✓ Branch 1 taken 27620 times.
+
✗ Branch 2 not taken.
+
+
+ 		27620 log_debug("New client thread thread %d has started\n", client->tid);
114 +
115 + /* Let the client thread handle its signals except those that should be
116 + * handled in the main thread. */
117 + 27620 ret = signal_block_sigterms();
118 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (ret < 0)
119 + ✗ goto cleanup;
120 +
121 + 27620 ret = client->server->conn_handler(client->conn_fd, client->server->conn_handler_arg);
122 +
+ 1/2 +
+
✓ Branch 0 taken 27620 times.
+
✗ Branch 1 not taken.
+
+
+ 		27620 if (ret < 0)
123 + ✗ goto cleanup;
124 +
125 + 27620cleanup:
126 +
+ 1/4 +
+
✗ Branch 1 not taken.
+
✓ Branch 2 taken 27620 times.
+
✗ Branch 4 not taken.
+
✗ Branch 5 not taken.
+
+
+ 		27620 log_errno_if(eventfd_write(client->cleanup_fd, 1), "eventfd_write");
127 +
128 + 27620 return NULL;
129 + }
130 +
131 + 27620static int client_create_thread(struct client *client)
132 + {
133 + sigset_t old_mask;
134 + 27620 int ret = 0;
135 +
136 + /* Block all signals (we'll unblock them later); the client thread will
137 + * have all signals blocked initially. This allows the main thread to
138 + * handle SIGINT/SIGTERM/etc. */
139 + 27620 ret = signal_block_all(&old_mask);
140 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (ret < 0)
141 + ✗ return ret;
142 +
143 + 27620 ret = pthread_create(&client->thread, NULL, client_thread_func, client);
144 +
+ 1/2 +
+
✓ Branch 0 taken 27620 times.
+
✗ Branch 1 not taken.
+
+
+ 		27620 if (ret) {
145 + ✗ pthread_errno(ret, "pthread_create");
146 + ✗ goto restore_mask;
147 + }
148 +
149 + 27620restore_mask:
150 + /* Restore the previously-enabled signals for handling in the main thread. */
151 + 27620 signal_set_mask(&old_mask);
152 +
153 + 27620 return ret;
154 + }
155 +
156 + 27620static int client_create(struct tcp_server *server, int conn_fd)
157 + {
158 + 27620 int ret = 0;
159 +
160 + 27620 struct client *client = calloc(1, sizeof(struct client));
161 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (!client) {
162 + ✗ log_errno("calloc");
163 + ✗ return -1;
164 + }
165 +
166 + 27620 client->server = server;
167 + 27620 client->conn_fd = conn_fd;
168 +
169 + 27620 ret = eventfd(0, EFD_CLOEXEC);
170 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (ret < 0) {
171 + ✗ log_errno("eventfd");
172 + ✗ goto free;
173 + }
174 + 27620 client->cleanup_fd = ret;
175 +
176 + 27620 ret = event_loop_add_once(server->loop, client->cleanup_fd, POLLIN, client_destroy_handler,
177 + client);
178 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (ret < 0)
179 + ✗ goto close_cleanup_fd;
180 +
181 + 27620 SIMPLEQ_INSERT_TAIL(&server->client_queue, client, entries);
182 +
183 + 27620 ret = client_create_thread(client);
184 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (ret < 0)
185 + ✗ goto remove_from_client_queue;
186 +
187 + 27620 return ret;
188 +
189 + ✗remove_from_client_queue:
190 + ✗ SIMPLEQ_REMOVE(&server->client_queue, client, client, entries);
191 +
192 + ✗close_cleanup_fd:
193 + ✗ file_close(client->cleanup_fd);
194 +
195 + ✗free:
196 + ✗ free(client);
197 +
198 + ✗ return ret;
199 + }
200 +
201 + 29static void client_queue_create(struct client_queue *client_queue)
202 + {
203 + 29 SIMPLEQ_INIT(client_queue);
204 + 29}
205 +
206 + 29static void client_queue_destroy(struct client_queue *client_queue)
207 + {
208 + 29 struct client *entry1 = SIMPLEQ_FIRST(client_queue);
209 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 29 times.
+
+
+ 		29 while (entry1) {
210 + ✗ struct client *entry2 = SIMPLEQ_NEXT(entry1, entries);
211 + ✗ client_destroy(entry1);
212 + ✗ entry1 = entry2;
213 + }
214 + 29}
215 +
216 + 27620static int tcp_server_accept_handler(UNUSED struct event_loop *loop, UNUSED int fd,
217 + UNUSED short revents, void *_server)
218 + {
219 + 27620 struct tcp_server *server = (struct tcp_server *)_server;
220 + 27620 return tcp_server_accept(server);
221 + }
222 +
223 + 29int tcp_server_create(struct tcp_server **_server, struct event_loop *loop, const char *port,
224 + tcp_server_conn_handler conn_handler, void *conn_handler_arg)
225 + {
226 + 29 int ret = 0;
227 +
228 + 29 struct tcp_server *server = calloc(1, sizeof(struct tcp_server));
229 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 29 times.
+
+
+ 		29 if (!server) {
230 + ✗ log_errno("calloc");
231 + ✗ return -1;
232 + }
233 +
234 + 29 server->loop = loop;
235 +
236 + 29 server->conn_handler = conn_handler;
237 + 29 server->conn_handler_arg = conn_handler_arg;
238 +
239 + 29 client_queue_create(&server->client_queue);
240 +
241 + 29 ret = net_bind(port);
242 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 29 times.
+
+
+ 		29 if (ret < 0)
243 + ✗ goto free;
244 + 29 server->accept_fd = ret;
245 +
246 + 29 ret = event_loop_add(loop, server->accept_fd, POLLIN, tcp_server_accept_handler, server);
247 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 29 times.
+
+
+ 		29 if (ret < 0)
248 + ✗ goto close;
249 +
250 + 29 *_server = server;
251 + 29 return ret;
252 +
253 + ✗close:
254 + ✗ net_close(server->accept_fd);
255 + ✗free:
256 + ✗ free(server);
257 +
258 + ✗ return ret;
259 + }
260 +
261 + 29void tcp_server_destroy(struct tcp_server *server)
262 + {
263 + 29 net_close(server->accept_fd);
264 + 29 client_queue_destroy(&server->client_queue);
265 + 29 free(server);
266 + 29}
267 +
268 + 27620int tcp_server_accept(struct tcp_server *server)
269 + {
270 + 27620 int conn_fd = -1, ret = 0;
271 +
272 + 27620 ret = net_accept(server->accept_fd);
273 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (ret < 0)
274 + ✗ return ret;
275 + 27620 conn_fd = ret;
276 +
277 + 27620 ret = client_create(server, conn_fd);
278 +
+ 1/2 +
+
✗ Branch 0 not taken.
+
✓ Branch 1 taken 27620 times.
+
+
+ 		27620 if (ret < 0)
279 + ✗ goto close_conn;
280 +
281 + 27620 return ret;
282 +
283 + ✗close_conn:
284 + ✗ net_close(conn_fd);
285 +
286 + ✗ return ret;
287 + }
288 +
+
+ +
+ + + + -- cgit v1.2.3