summaryrefslogtreecommitdiff
path: root/net-misc/curl/files/curl-8.1.0-numeric-hostname.patch
blob: 6a0dd1382d62de62abb18f661c28b7bbd8f5dbad (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
https://github.com/curl/curl/commit/92772e6d395bbdda0e7822d980caf86e8c4aa51c.patch
From: Daniel Stenberg <daniel@haxx.se>
Date: Thu, 18 May 2023 00:31:17 +0200
Subject: [PATCH] urlapi: allow numerical parts in the host name

It can only be an IPv4 address if all parts are all digits and no more than
four parts, otherwise it is a host name. Even slightly wrong IPv4 will now be
passed through as a host name.

Regression from 17a15d88467 shipped in 8.1.0

Extended test 1560 accordingly.

Reported-by: Pavel Kalyugin
Fixes #11129
Closes #11131
--- a/lib/urlapi.c
+++ b/lib/urlapi.c
@@ -34,6 +34,7 @@
 #include "inet_ntop.h"
 #include "strdup.h"
 #include "idn.h"
+#include "curl_memrchr.h"
 
 /* The last 3 #include files should be in this order */
 #include "curl_printf.h"
@@ -643,8 +644,8 @@ static CURLUcode hostname_check(struct Curl_URL *u, char *hostname,
  * Handle partial IPv4 numerical addresses and different bases, like
  * '16843009', '0x7f', '0x7f.1' '0177.1.1.1' etc.
  *
- * If the given input string is syntactically wrong or any part for example is
- * too big, this function returns FALSE and doesn't create any output.
+ * If the given input string is syntactically wrong IPv4 or any part for
+ * example is too big, this function returns HOST_NAME.
  *
  * Output the "normalized" version of that input string in plain quad decimal
  * integers.
@@ -675,7 +676,7 @@ static int ipv4_normalize(struct dynbuf *host)
     unsigned long l;
     if(!ISDIGIT(*c))
       /* most importantly this doesn't allow a leading plus or minus */
-      return n ? HOST_BAD : HOST_NAME;
+      return HOST_NAME;
     l = strtoul(c, &endp, 0);
 
     parts[n] = l;
@@ -684,7 +685,7 @@ static int ipv4_normalize(struct dynbuf *host)
     switch(*c) {
     case '.':
       if(n == 3)
-        return HOST_BAD;
+        return HOST_NAME;
       n++;
       c++;
       break;
@@ -694,39 +695,40 @@ static int ipv4_normalize(struct dynbuf *host)
       break;
 
     default:
-      return n ? HOST_BAD : HOST_NAME;
+      return HOST_NAME;
     }
 
     /* overflow */
     if((l == ULONG_MAX) && (errno == ERANGE))
-      return HOST_BAD;
+      return HOST_NAME;
 
 #if SIZEOF_LONG > 4
     /* a value larger than 32 bits */
     if(l > UINT_MAX)
-      return HOST_BAD;
+      return HOST_NAME;
 #endif
   }
 
-  /* this is a valid IPv4 numerical address */
-  Curl_dyn_reset(host);
-
   switch(n) {
   case 0: /* a -- 32 bits */
+    Curl_dyn_reset(host);
+
     result = Curl_dyn_addf(host, "%u.%u.%u.%u",
                            parts[0] >> 24, (parts[0] >> 16) & 0xff,
                            (parts[0] >> 8) & 0xff, parts[0] & 0xff);
     break;
   case 1: /* a.b -- 8.24 bits */
     if((parts[0] > 0xff) || (parts[1] > 0xffffff))
-      return HOST_BAD;
+      return HOST_NAME;
+    Curl_dyn_reset(host);
     result = Curl_dyn_addf(host, "%u.%u.%u.%u",
                            parts[0], (parts[1] >> 16) & 0xff,
                            (parts[1] >> 8) & 0xff, parts[1] & 0xff);
     break;
   case 2: /* a.b.c -- 8.8.16 bits */
     if((parts[0] > 0xff) || (parts[1] > 0xff) || (parts[2] > 0xffff))
-      return HOST_BAD;
+      return HOST_NAME;
+    Curl_dyn_reset(host);
     result = Curl_dyn_addf(host, "%u.%u.%u.%u",
                            parts[0], parts[1], (parts[2] >> 8) & 0xff,
                            parts[2] & 0xff);
@@ -734,7 +736,8 @@ static int ipv4_normalize(struct dynbuf *host)
   case 3: /* a.b.c.d -- 8.8.8.8 bits */
     if((parts[0] > 0xff) || (parts[1] > 0xff) || (parts[2] > 0xff) ||
        (parts[3] > 0xff))
-      return HOST_BAD;
+      return HOST_NAME;
+    Curl_dyn_reset(host);
     result = Curl_dyn_addf(host, "%u.%u.%u.%u",
                            parts[0], parts[1], parts[2], parts[3]);
     break;
@@ -796,6 +799,9 @@ static CURLUcode parse_authority(struct Curl_URL *u,
   if(result)
     goto out;
 
+  if(!Curl_dyn_len(host))
+    return CURLUE_NO_HOST;
+
   switch(ipv4_normalize(host)) {
   case HOST_IPV4:
     break;
--- a/tests/libtest/lib1560.c
+++ b/tests/libtest/lib1560.c
@@ -474,6 +474,13 @@ static const struct testcase get_parts_list[] ={
 };
 
 static const struct urltestcase get_url_list[] = {
+  {"https://1.0x1000000", "https://1.0x1000000/", 0, 0, CURLUE_OK},
+  {"https://0x7f.1", "https://127.0.0.1/", 0, 0, CURLUE_OK},
+  {"https://1.2.3.256.com", "https://1.2.3.256.com/", 0, 0, CURLUE_OK},
+  {"https://10.com", "https://10.com/", 0, 0, CURLUE_OK},
+  {"https://1.2.com", "https://1.2.com/", 0, 0, CURLUE_OK},
+  {"https://1.2.3.com", "https://1.2.3.com/", 0, 0, CURLUE_OK},
+  {"https://1.2.com.99", "https://1.2.com.99/", 0, 0, CURLUE_OK},
   {"https://[fe80::0000:20c:29ff:fe9c:409b]:80/moo",
    "https://[fe80::20c:29ff:fe9c:409b]:80/moo",
    0, 0, CURLUE_OK},
@@ -522,22 +529,24 @@ static const struct urltestcase get_url_list[] = {
 
   /* IPv4 trickeries */
   {"https://16843009", "https://1.1.1.1/", 0, 0, CURLUE_OK},
-  {"https://0x7f.1", "https://127.0.0.1/", 0, 0, CURLUE_OK},
   {"https://0177.1", "https://127.0.0.1/", 0, 0, CURLUE_OK},
   {"https://0111.02.0x3", "https://73.2.0.3/", 0, 0, CURLUE_OK},
+  {"https://0111.02.0x3.", "https://0111.02.0x3./", 0, 0, CURLUE_OK},
+  {"https://0111.02.030", "https://73.2.0.24/", 0, 0, CURLUE_OK},
+  {"https://0111.02.030.", "https://0111.02.030./", 0, 0, CURLUE_OK},
   {"https://0xff.0xff.0377.255", "https://255.255.255.255/", 0, 0, CURLUE_OK},
   {"https://1.0xffffff", "https://1.255.255.255/", 0, 0, CURLUE_OK},
   /* IPv4 numerical overflows or syntax errors will not normalize */
   {"https://a127.0.0.1", "https://a127.0.0.1/", 0, 0, CURLUE_OK},
   {"https://\xff.127.0.0.1", "https://%FF.127.0.0.1/", 0, CURLU_URLENCODE,
    CURLUE_OK},
-  {"https://127.-0.0.1", "https://127.-0.0.1/", 0, 0, CURLUE_BAD_HOSTNAME},
+  {"https://127.-0.0.1", "https://127.-0.0.1/", 0, 0, CURLUE_OK},
   {"https://127.0. 1", "https://127.0.0.1/", 0, 0, CURLUE_MALFORMED_INPUT},
-  {"https://1.0x1000000", "https://1.0x1000000/", 0, 0, CURLUE_BAD_HOSTNAME},
-  {"https://1.2.3.256", "https://1.2.3.256/", 0, 0, CURLUE_BAD_HOSTNAME},
-  {"https://1.2.3.4.5", "https://1.2.3.4.5/", 0, 0, CURLUE_BAD_HOSTNAME},
-  {"https://1.2.0x100.3", "https://1.2.0x100.3/", 0, 0, CURLUE_BAD_HOSTNAME},
-  {"https://4294967296", "https://4294967296/", 0, 0, CURLUE_BAD_HOSTNAME},
+  {"https://1.2.3.256", "https://1.2.3.256/", 0, 0, CURLUE_OK},
+  {"https://1.2.3.256.", "https://1.2.3.256./", 0, 0, CURLUE_OK},
+  {"https://1.2.3.4.5", "https://1.2.3.4.5/", 0, 0, CURLUE_OK},
+  {"https://1.2.0x100.3", "https://1.2.0x100.3/", 0, 0, CURLUE_OK},
+  {"https://4294967296", "https://4294967296/", 0, 0, CURLUE_OK},
   {"https://123host", "https://123host/", 0, 0, CURLUE_OK},
   /* 40 bytes scheme is the max allowed */
   {"AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA://hostname/path",
@@ -599,20 +608,11 @@ static const struct urltestcase get_url_list[] = {
    0, 0, CURLUE_OK},
   /* here the password has the semicolon */
   {"http://user:pass;word@host/file",
-   "http://user:pass;word@host/file",
-   0, 0, CURLUE_OK},
-  {"file:///file.txt#moo",
-   "file:///file.txt#moo",
-   0, 0, CURLUE_OK},
-  {"file:////file.txt",
-   "file:////file.txt",
-   0, 0, CURLUE_OK},
-  {"file:///file.txt",
-   "file:///file.txt",
-   0, 0, CURLUE_OK},
-  {"file:./",
-   "file://",
-   0, 0, CURLUE_BAD_SCHEME},
+   "http://user:pass;word@host/file", 0, 0, CURLUE_OK},
+  {"file:///file.txt#moo", "file:///file.txt#moo", 0, 0, CURLUE_OK},
+  {"file:////file.txt", "file:////file.txt", 0, 0, CURLUE_OK},
+  {"file:///file.txt", "file:///file.txt", 0, 0, CURLUE_OK},
+  {"file:./", "file://", 0, 0, CURLUE_OK},
   {"http://example.com/hello/../here",
    "http://example.com/hello/../here",
    CURLU_PATH_AS_IS, 0, CURLUE_OK},
@@ -1124,7 +1124,7 @@ static int get_url(void)
       }
       curl_free(url);
     }
-    else if(rc != get_url_list[i].ucode) {
+    if(rc != get_url_list[i].ucode) {
       fprintf(stderr, "Get URL\nin: %s\nreturned %d (expected %d)\n",
               get_url_list[i].in, (int)rc, get_url_list[i].ucode);
       error++;
@@ -1515,6 +1515,9 @@ int test(char *URL)
 {
   (void)URL; /* not used */
 
+  if(get_url())
+    return 3;
+
   if(huge())
     return 9;
 
@@ -1533,9 +1536,6 @@ int test(char *URL)
   if(set_parts())
     return 2;
 
-  if(get_url())
-    return 3;
-
   if(get_parts())
     return 4;