| 1 |
.\" -*- mode: troff; coding: utf-8 -*- |
| 2 |
.\" Automatically generated by Pod::Man 5.0102 (Pod::Simple 3.45) |
| 3 |
.\" |
| 4 |
.\" Standard preamble: |
| 5 |
.\" ======================================================================== |
| 6 |
.de Sp \" Vertical space (when we can't use .PP) |
| 7 |
.if t .sp .5v |
| 8 |
.if n .sp |
| 9 |
.. |
| 10 |
.de Vb \" Begin verbatim text |
| 11 |
.ft CW |
| 12 |
.nf |
| 13 |
.ne \\$1 |
| 14 |
.. |
| 15 |
.de Ve \" End verbatim text |
| 16 |
.ft R |
| 17 |
.fi |
| 18 |
.. |
| 19 |
.\" \*(C` and \*(C' are quotes in nroff, nothing in troff, for use with C<>. |
| 20 |
.ie n \{\ |
| 21 |
. ds C` "" |
| 22 |
. ds C' "" |
| 23 |
'br\} |
| 24 |
.el\{\ |
| 25 |
. ds C` |
| 26 |
. ds C' |
| 27 |
'br\} |
| 28 |
.\" |
| 29 |
.\" Escape single quotes in literal strings from groff's Unicode transform. |
| 30 |
.ie \n(.g .ds Aq \(aq |
| 31 |
.el .ds Aq ' |
| 32 |
.\" |
| 33 |
.\" If the F register is >0, we'll generate index entries on stderr for |
| 34 |
.\" titles (.TH), headers (.SH), subsections (.SS), items (.Ip), and index |
| 35 |
.\" entries marked with X<> in POD. Of course, you'll have to process the |
| 36 |
.\" output yourself in some meaningful fashion. |
| 37 |
.\" |
| 38 |
.\" Avoid warning from groff about undefined register 'F'. |
| 39 |
.de IX |
| 40 |
.. |
| 41 |
.nr rF 0 |
| 42 |
.if \n(.g .if rF .nr rF 1 |
| 43 |
.if (\n(rF:(\n(.g==0)) \{\ |
| 44 |
. if \nF \{\ |
| 45 |
. de IX |
| 46 |
. tm Index:\\$1\t\\n%\t"\\$2" |
| 47 |
.. |
| 48 |
. if !\nF==2 \{\ |
| 49 |
. nr % 0 |
| 50 |
. nr F 2 |
| 51 |
. \} |
| 52 |
. \} |
| 53 |
.\} |
| 54 |
.rr rF |
| 55 |
.\" ======================================================================== |
| 56 |
.\" |
| 57 |
.IX Title "OSSL-GUIDE-QUIC-CLIENT-NON-BLOCK 7SSL" |
| 58 |
.TH OSSL-GUIDE-QUIC-CLIENT-NON-BLOCK 7SSL 2026-04-07 3.6.2 OpenSSL |
| 59 |
.\" For nroff, turn off justification. Always turn off hyphenation; it makes |
| 60 |
.\" way too many mistakes in technical documents. |
| 61 |
.if n .ad l |
| 62 |
.nh |
| 63 |
.SH NAME |
| 64 |
ossl\-guide\-quic\-client\-non\-block |
| 65 |
\&\- OpenSSL Guide: Writing a simple nonblocking QUIC client |
| 66 |
.SH "SIMPLE NONBLOCKING QUIC CLIENT EXAMPLE" |
| 67 |
.IX Header "SIMPLE NONBLOCKING QUIC CLIENT EXAMPLE" |
| 68 |
This page will build on the example developed on the |
| 69 |
\&\fBossl\-guide\-quic\-client\-block\fR\|(7) page which demonstrates how to write a simple |
| 70 |
blocking QUIC client. On this page we will amend that demo code so that it |
| 71 |
supports nonblocking functionality. |
| 72 |
.PP |
| 73 |
The complete source code for this example nonblocking QUIC client is available |
| 74 |
in the \fBdemos/guide\fR directory of the OpenSSL source distribution in the file |
| 75 |
\&\fBquic\-client\-non\-block.c\fR. It is also available online at |
| 76 |
<https://github.com/openssl/openssl/blob/master/demos/guide/quic\-client\-non\-block.c>. |
| 77 |
.PP |
| 78 |
As we saw in the previous example an OpenSSL QUIC application always uses a |
| 79 |
nonblocking socket. However, despite this, the \fBSSL\fR object still has blocking |
| 80 |
behaviour. When the \fBSSL\fR object has blocking behaviour then this means that |
| 81 |
it waits (blocks) until data is available to read if you attempt to read from |
| 82 |
it when there is no data yet. Similarly it waits when writing if the \fBSSL\fR |
| 83 |
object is currently unable to write at the moment. This can simplify the |
| 84 |
development of code because you do not have to worry about what to do in these |
| 85 |
cases. The execution of the code will simply stop until it is able to continue. |
| 86 |
However in many cases you do not want this behaviour. Rather than stopping and |
| 87 |
waiting your application may need to go and do other tasks whilst the \fBSSL\fR |
| 88 |
object is unable to read/write, for example updating a GUI or performing |
| 89 |
operations on some other connection or stream. |
| 90 |
.PP |
| 91 |
We will see later in this tutorial how to change the \fBSSL\fR object so that it |
| 92 |
has nonblocking behaviour. With a nonblocking \fBSSL\fR object, functions such as |
| 93 |
\&\fBSSL_read_ex\fR\|(3) or \fBSSL_write_ex\fR\|(3) will return immediately with a non-fatal |
| 94 |
error if they are currently unable to read or write respectively. |
| 95 |
.PP |
| 96 |
Since this page is building on the example developed on the |
| 97 |
\&\fBossl\-guide\-quic\-client\-block\fR\|(7) page we assume that you are familiar with it |
| 98 |
and we only explain how this example differs. |
| 99 |
.SS "Performing work while waiting for the socket" |
| 100 |
.IX Subsection "Performing work while waiting for the socket" |
| 101 |
In a nonblocking application you will need work to perform in the event that |
| 102 |
we want to read or write to the \fBSSL\fR object but we are currently unable to. |
| 103 |
In fact this is the whole point of using a nonblocking \fBSSL\fR object, i.e. to |
| 104 |
give the application the opportunity to do something else. Whatever it is that |
| 105 |
the application has to do, it must also be prepared to come back and retry the |
| 106 |
operation that it previously attempted periodically to see if it can now |
| 107 |
complete. Ideally it would only do this in the event that something has changed |
| 108 |
such that it might succeed on the retry attempt, but this does not have to be |
| 109 |
the case. It can retry at any time. |
| 110 |
.PP |
| 111 |
Note that it is important that you retry exactly the same operation that you |
| 112 |
tried last time. You cannot start something new. For example if you were |
| 113 |
attempting to write the text "Hello World" and the operation failed because the |
| 114 |
\&\fBSSL\fR object is currently unable to write, then you cannot then attempt to |
| 115 |
write some other text when you retry the operation. |
| 116 |
.PP |
| 117 |
In this demo application we will create a helper function which simulates doing |
| 118 |
other work. In fact, for the sake of simplicity, it will do nothing except wait |
| 119 |
for the state of the underlying socket to change or until a timeout expires |
| 120 |
after which the state of the \fBSSL\fR object might have changed. We will call our |
| 121 |
function \f(CWwait_for_activity()\fR. |
| 122 |
.PP |
| 123 |
.Vb 6 |
| 124 |
\& static void wait_for_activity(SSL *ssl) |
| 125 |
\& { |
| 126 |
\& fd_set wfds, rfds; |
| 127 |
\& int width, sock, isinfinite; |
| 128 |
\& struct timeval tv; |
| 129 |
\& struct timeval *tvp = NULL; |
| 130 |
\& |
| 131 |
\& /* Get hold of the underlying file descriptor for the socket */ |
| 132 |
\& sock = SSL_get_fd(ssl); |
| 133 |
\& |
| 134 |
\& FD_ZERO(&wfds); |
| 135 |
\& FD_ZERO(&rfds); |
| 136 |
\& |
| 137 |
\& /* |
| 138 |
\& * Find out if we would like to write to the socket, or read from it (or |
| 139 |
\& * both) |
| 140 |
\& */ |
| 141 |
\& if (SSL_net_write_desired(ssl)) |
| 142 |
\& FD_SET(sock, &wfds); |
| 143 |
\& if (SSL_net_read_desired(ssl)) |
| 144 |
\& FD_SET(sock, &rfds); |
| 145 |
\& width = sock + 1; |
| 146 |
\& |
| 147 |
\& /* |
| 148 |
\& * Find out when OpenSSL would next like to be called, regardless of |
| 149 |
\& * whether the state of the underlying socket has changed or not. |
| 150 |
\& */ |
| 151 |
\& if (SSL_get_event_timeout(ssl, &tv, &isinfinite) && !isinfinite) |
| 152 |
\& tvp = &tv; |
| 153 |
\& |
| 154 |
\& /* |
| 155 |
\& * Wait until the socket is writeable or readable. We use select here |
| 156 |
\& * for the sake of simplicity and portability, but you could equally use |
| 157 |
\& * poll/epoll or similar functions |
| 158 |
\& * |
| 159 |
\& * NOTE: For the purposes of this demonstration code this effectively |
| 160 |
\& * makes this demo block until it has something more useful to do. In a |
| 161 |
\& * real application you probably want to go and do other work here (e.g. |
| 162 |
\& * update a GUI, or service other connections). |
| 163 |
\& * |
| 164 |
\& * Let\*(Aqs say for example that you want to update the progress counter on |
| 165 |
\& * a GUI every 100ms. One way to do that would be to use the timeout in |
| 166 |
\& * the last parameter to "select" below. If the tvp value is greater |
| 167 |
\& * than 100ms then use 100ms instead. Then, when select returns, you |
| 168 |
\& * check if it did so because of activity on the file descriptors or |
| 169 |
\& * because of the timeout. If the 100ms GUI timeout has expired but the |
| 170 |
\& * tvp timeout has not then go and update the GUI and then restart the |
| 171 |
\& * "select" (with updated timeouts). |
| 172 |
\& */ |
| 173 |
\& |
| 174 |
\& select(width, &rfds, &wfds, NULL, tvp); |
| 175 |
\&} |
| 176 |
.Ve |
| 177 |
.PP |
| 178 |
If you are familiar with how to write nonblocking applications in OpenSSL for |
| 179 |
TLS (see \fBossl\-guide\-tls\-client\-non\-block\fR\|(7)) then you should note that there |
| 180 |
is an important difference here between the way a QUIC application and a TLS |
| 181 |
application works. With a TLS application if we try to read or write something |
| 182 |
to the \fBSSL\fR object and we get a "retry" response (\fBSSL_ERROR_WANT_READ\fR or |
| 183 |
\&\fBSSL_ERROR_WANT_WRITE\fR) then we can assume that is because OpenSSL attempted to |
| 184 |
read or write to the underlying socket and the socket signalled the "retry". |
| 185 |
With QUIC that is not the case. OpenSSL may signal retry as a result of an |
| 186 |
\&\fBSSL_read_ex\fR\|(3) or \fBSSL_write_ex\fR\|(3) (or similar) call which indicates the |
| 187 |
state of the stream. This is entirely independent of whether the underlying |
| 188 |
socket needs to retry or not. |
| 189 |
.PP |
| 190 |
To determine whether OpenSSL currently wants to read or write to the underlying |
| 191 |
socket for a QUIC application we must call the \fBSSL_net_read_desired\fR\|(3) and |
| 192 |
\&\fBSSL_net_write_desired\fR\|(3) functions. |
| 193 |
.PP |
| 194 |
It is also important with QUIC that we periodically call an I/O function (or |
| 195 |
otherwise call the \fBSSL_handle_events\fR\|(3) function) to ensure that the QUIC |
| 196 |
connection remains healthy. This is particularly important with a nonblocking |
| 197 |
application because you are likely to leave the \fBSSL\fR object idle for a while |
| 198 |
while the application goes off to do other work. The \fBSSL_get_event_timeout\fR\|(3) |
| 199 |
function can be used to determine what the deadline is for the next time we need |
| 200 |
to call an I/O function (or call \fBSSL_handle_events\fR\|(3)). |
| 201 |
.PP |
| 202 |
An alternative to using \fBSSL_get_event_timeout\fR\|(3) to find the next deadline |
| 203 |
that OpenSSL must be called again by is to use "thread assisted" mode. In |
| 204 |
"thread assisted" mode OpenSSL spawns an additional thread which will |
| 205 |
periodically call \fBSSL_handle_events\fR\|(3) automatically, meaning that the |
| 206 |
application can leave the connection idle safe in the knowledge that the |
| 207 |
connection will still be maintained in a healthy state. See |
| 208 |
"Creating the SSL_CTX and SSL objects" below for further details about this. |
| 209 |
.PP |
| 210 |
In this example we are using the \f(CW\*(C`select\*(C'\fR function to check the |
| 211 |
readability/writeability of the socket because it is very simple to use and is |
| 212 |
available on most Operating Systems. However you could use any other similar |
| 213 |
function to do the same thing. \f(CW\*(C`select\*(C'\fR waits for the state of the underlying |
| 214 |
socket(s) to become readable/writeable or until the timeout has expired before |
| 215 |
returning. |
| 216 |
.SS "Handling errors from OpenSSL I/O functions" |
| 217 |
.IX Subsection "Handling errors from OpenSSL I/O functions" |
| 218 |
A QUIC application that has been configured for nonblocking behaviour will need |
| 219 |
to be prepared to handle errors returned from OpenSSL I/O functions such as |
| 220 |
\&\fBSSL_read_ex\fR\|(3) or \fBSSL_write_ex\fR\|(3). Errors may be fatal for the stream (for |
| 221 |
example because the stream has been reset or because the underlying connection |
| 222 |
has failed), or non-fatal (for example because we are trying to read from the |
| 223 |
stream but no data has not yet arrived from the peer for that stream). |
| 224 |
.PP |
| 225 |
\&\fBSSL_read_ex\fR\|(3) and \fBSSL_write_ex\fR\|(3) will return 0 to indicate an error and |
| 226 |
\&\fBSSL_read\fR\|(3) and \fBSSL_write\fR\|(3) will return 0 or a negative value to indicate |
| 227 |
an error. \fBSSL_shutdown\fR\|(3) will return a negative value to incidate an error. |
| 228 |
.PP |
| 229 |
In the event of an error an application should call \fBSSL_get_error\fR\|(3) to find |
| 230 |
out what type of error has occurred. If the error is non-fatal and can be |
| 231 |
retried then \fBSSL_get_error\fR\|(3) will return \fBSSL_ERROR_WANT_READ\fR or |
| 232 |
\&\fBSSL_ERROR_WANT_WRITE\fR depending on whether OpenSSL wanted to read to or write |
| 233 |
from the stream but was unable to. Note that a call to \fBSSL_read_ex\fR\|(3) or |
| 234 |
\&\fBSSL_read\fR\|(3) can still generate \fBSSL_ERROR_WANT_WRITE\fR. Similarly calls to |
| 235 |
\&\fBSSL_write_ex\fR\|(3) or \fBSSL_write\fR\|(3) might generate \fBSSL_ERROR_WANT_READ\fR. |
| 236 |
.PP |
| 237 |
Another type of non-fatal error that may occur is \fBSSL_ERROR_ZERO_RETURN\fR. This |
| 238 |
indicates an EOF (End-Of-File) which can occur if you attempt to read data from |
| 239 |
an \fBSSL\fR object but the peer has indicated that it will not send any more data |
| 240 |
on the stream. In this case you may still want to write data to the stream but |
| 241 |
you will not receive any more data. |
| 242 |
.PP |
| 243 |
Fatal errors that may occur are \fBSSL_ERROR_SYSCALL\fR and \fBSSL_ERROR_SSL\fR. These |
| 244 |
indicate that the stream is no longer usable. For example, this could be because |
| 245 |
the stream has been reset by the peer, or because the underlying connection has |
| 246 |
failed. You can consult the OpenSSL error stack for further details (for example |
| 247 |
by calling \fBERR_print_errors\fR\|(3) to print out details of errors that have |
| 248 |
occurred). You can also consult the return value of |
| 249 |
\&\fBSSL_get_stream_read_state\fR\|(3) to determine whether the error is local to the |
| 250 |
stream, or whether the underlying connection has also failed. A return value |
| 251 |
of \fBSSL_STREAM_STATE_RESET_REMOTE\fR tells you that the stream has been reset by |
| 252 |
the peer and \fBSSL_STREAM_STATE_CONN_CLOSED\fR tells you that the underlying |
| 253 |
connection has closed. |
| 254 |
.PP |
| 255 |
In our demo application we will write a function to handle these errors from |
| 256 |
OpenSSL I/O functions: |
| 257 |
.PP |
| 258 |
.Vb 8 |
| 259 |
\& static int handle_io_failure(SSL *ssl, int res) |
| 260 |
\& { |
| 261 |
\& switch (SSL_get_error(ssl, res)) { |
| 262 |
\& case SSL_ERROR_WANT_READ: |
| 263 |
\& case SSL_ERROR_WANT_WRITE: |
| 264 |
\& /* Temporary failure. Wait until we can read/write and try again */ |
| 265 |
\& wait_for_activity(ssl); |
| 266 |
\& return 1; |
| 267 |
\& |
| 268 |
\& case SSL_ERROR_ZERO_RETURN: |
| 269 |
\& /* EOF */ |
| 270 |
\& return 0; |
| 271 |
\& |
| 272 |
\& case SSL_ERROR_SYSCALL: |
| 273 |
\& return \-1; |
| 274 |
\& |
| 275 |
\& case SSL_ERROR_SSL: |
| 276 |
\& /* |
| 277 |
\& * Some stream fatal error occurred. This could be because of a |
| 278 |
\& * stream reset \- or some failure occurred on the underlying |
| 279 |
\& * connection. |
| 280 |
\& */ |
| 281 |
\& switch (SSL_get_stream_read_state(ssl)) { |
| 282 |
\& case SSL_STREAM_STATE_RESET_REMOTE: |
| 283 |
\& printf("Stream reset occurred\en"); |
| 284 |
\& /* |
| 285 |
\& * The stream has been reset but the connection is still |
| 286 |
\& * healthy. |
| 287 |
\& */ |
| 288 |
\& break; |
| 289 |
\& |
| 290 |
\& case SSL_STREAM_STATE_CONN_CLOSED: |
| 291 |
\& printf("Connection closed\en"); |
| 292 |
\& /* Connection is already closed. */ |
| 293 |
\& break; |
| 294 |
\& |
| 295 |
\& default: |
| 296 |
\& printf("Unknown stream failure\en"); |
| 297 |
\& break; |
| 298 |
\& } |
| 299 |
\& /* |
| 300 |
\& * If the failure is due to a verification error we can get more |
| 301 |
\& * information about it from SSL_get_verify_result(). |
| 302 |
\& */ |
| 303 |
\& if (SSL_get_verify_result(ssl) != X509_V_OK) |
| 304 |
\& printf("Verify error: %s\en", |
| 305 |
\& X509_verify_cert_error_string(SSL_get_verify_result(ssl))); |
| 306 |
\& return \-1; |
| 307 |
\& |
| 308 |
\& default: |
| 309 |
\& return \-1; |
| 310 |
\& } |
| 311 |
\& } |
| 312 |
.Ve |
| 313 |
.PP |
| 314 |
This function takes as arguments the \fBSSL\fR object that represents the |
| 315 |
connection, as well as the return code from the I/O function that failed. In |
| 316 |
the event of a non-fatal failure, it waits until a retry of the I/O operation |
| 317 |
might succeed (by using the \f(CWwait_for_activity()\fR function that we developed |
| 318 |
in the previous section). It returns 1 in the event of a non-fatal error |
| 319 |
(except EOF), 0 in the event of EOF, or \-1 if a fatal error occurred. |
| 320 |
.SS "Creating the SSL_CTX and SSL objects" |
| 321 |
.IX Subsection "Creating the SSL_CTX and SSL objects" |
| 322 |
In order to connect to a server we must create \fBSSL_CTX\fR and \fBSSL\fR objects for |
| 323 |
this. Most of the steps to do this are the same as for a blocking client and are |
| 324 |
explained on the \fBossl\-guide\-quic\-client\-block\fR\|(7) page. We won't repeat that |
| 325 |
information here. |
| 326 |
.PP |
| 327 |
One key difference is that we must put the \fBSSL\fR object into nonblocking mode |
| 328 |
(the default is blocking mode). To do that we use the |
| 329 |
\&\fBSSL_set_blocking_mode\fR\|(3) function: |
| 330 |
.PP |
| 331 |
.Vb 9 |
| 332 |
\& /* |
| 333 |
\& * The underlying socket is always nonblocking with QUIC, but the default |
| 334 |
\& * behaviour of the SSL object is still to block. We set it for nonblocking |
| 335 |
\& * mode in this demo. |
| 336 |
\& */ |
| 337 |
\& if (!SSL_set_blocking_mode(ssl, 0)) { |
| 338 |
\& printf("Failed to turn off blocking mode\en"); |
| 339 |
\& goto end; |
| 340 |
\& } |
| 341 |
.Ve |
| 342 |
.PP |
| 343 |
Although the demo application that we are developing here does not use it, it is |
| 344 |
possible to use "thread assisted mode" when developing QUIC applications. |
| 345 |
Normally, when writing an OpenSSL QUIC application, it is important that |
| 346 |
\&\fBSSL_handle_events\fR\|(3) (or alternatively any I/O function) is called on the |
| 347 |
connection \fBSSL\fR object periodically to maintain the connection in a healthy |
| 348 |
state. See "Performing work while waiting for the socket" for more discussion |
| 349 |
on this. This is particularly important to keep in mind when writing a |
| 350 |
nonblocking QUIC application because it is common to leave the \fBSSL\fR connection |
| 351 |
object idle for some time when using nonblocking mode. By using "thread assisted |
| 352 |
mode" a separate thread is created by OpenSSL to do this automatically which |
| 353 |
means that the application developer does not need to handle this aspect. To do |
| 354 |
this we must use \fBOSSL_QUIC_client_thread_method\fR\|(3) when we construct the |
| 355 |
\&\fBSSL_CTX\fR as shown below: |
| 356 |
.PP |
| 357 |
.Vb 5 |
| 358 |
\& ctx = SSL_CTX_new(OSSL_QUIC_client_thread_method()); |
| 359 |
\& if (ctx == NULL) { |
| 360 |
\& printf("Failed to create the SSL_CTX\en"); |
| 361 |
\& goto end; |
| 362 |
\& } |
| 363 |
.Ve |
| 364 |
.SS "Performing the handshake" |
| 365 |
.IX Subsection "Performing the handshake" |
| 366 |
As in the demo for a blocking QUIC client we use the \fBSSL_connect\fR\|(3) function |
| 367 |
to perform the handshake with the server. Since we are using a nonblocking |
| 368 |
\&\fBSSL\fR object it is very likely that calls to this function will fail with a |
| 369 |
non-fatal error while we are waiting for the server to respond to our handshake |
| 370 |
messages. In such a case we must retry the same \fBSSL_connect\fR\|(3) call at a |
| 371 |
later time. In this demo we do this in a loop: |
| 372 |
.PP |
| 373 |
.Vb 7 |
| 374 |
\& /* Do the handshake with the server */ |
| 375 |
\& while ((ret = SSL_connect(ssl)) != 1) { |
| 376 |
\& if (handle_io_failure(ssl, ret) == 1) |
| 377 |
\& continue; /* Retry */ |
| 378 |
\& printf("Failed to connect to server\en"); |
| 379 |
\& goto end; /* Cannot retry: error */ |
| 380 |
\& } |
| 381 |
.Ve |
| 382 |
.PP |
| 383 |
We continually call \fBSSL_connect\fR\|(3) until it gives us a success response. |
| 384 |
Otherwise we use the \f(CWhandle_io_failure()\fR function that we created earlier to |
| 385 |
work out what we should do next. Note that we do not expect an EOF to occur at |
| 386 |
this stage, so such a response is treated in the same way as a fatal error. |
| 387 |
.SS "Sending and receiving data" |
| 388 |
.IX Subsection "Sending and receiving data" |
| 389 |
As with the blocking QUIC client demo we use the \fBSSL_write_ex\fR\|(3) function to |
| 390 |
send data to the server. As with \fBSSL_connect\fR\|(3) above, because we are using |
| 391 |
a nonblocking \fBSSL\fR object, this call could fail with a non-fatal error. In |
| 392 |
that case we should retry exactly the same \fBSSL_write_ex\fR\|(3) call again. Note |
| 393 |
that the parameters must be \fIexactly\fR the same, i.e. the same pointer to the |
| 394 |
buffer to write with the same length. You must not attempt to send different |
| 395 |
data on a retry. An optional mode does exist |
| 396 |
(\fBSSL_MODE_ACCEPT_MOVING_WRITE_BUFFER\fR) which will configure OpenSSL to allow |
| 397 |
the buffer being written to change from one retry to the next. However, in this |
| 398 |
case, you must still retry exactly the same data \- even though the buffer that |
| 399 |
contains that data may change location. See \fBSSL_CTX_set_mode\fR\|(3) for further |
| 400 |
details. As in the TLS tutorials (\fBossl\-guide\-tls\-client\-block\fR\|(7)) we write |
| 401 |
the request in three chunks. |
| 402 |
.PP |
| 403 |
First, we write the entire request to the stream. We also must make sure to |
| 404 |
signal to the server that we have finished writing. This can be done by passing |
| 405 |
the SSL_WRITE_FLAG_CONCLUDE flag to \fBSSL_write_ex2\fR\|(3) or by calling |
| 406 |
\&\fBSSL_stream_conclude\fR\|(3). Since the first way is more efficient, we choose to |
| 407 |
do that. |
| 408 |
.PP |
| 409 |
.Vb 10 |
| 410 |
\& /* Write an HTTP GET request to the peer */ |
| 411 |
\& while (!SSL_write_ex(ssl, request_start, strlen(request_start), &written)) { |
| 412 |
\& if (handle_io_failure(ssl, 0) == 1) |
| 413 |
\& continue; /* Retry */ |
| 414 |
\& printf("Failed to write start of HTTP request\en"); |
| 415 |
\& goto end; /* Cannot retry: error */ |
| 416 |
\& } |
| 417 |
\& while (!SSL_write_ex(ssl, hostname, strlen(hostname), &written)) { |
| 418 |
\& if (handle_io_failure(ssl, 0) == 1) |
| 419 |
\& continue; /* Retry */ |
| 420 |
\& printf("Failed to write hostname in HTTP request\en"); |
| 421 |
\& goto end; /* Cannot retry: error */ |
| 422 |
\& } |
| 423 |
\& while (!SSL_write_ex2(ssl, request_end, strlen(request_end), |
| 424 |
\& SSL_WRITE_FLAG_CONCLUDE, &written)) { |
| 425 |
\& if (handle_io_failure(ssl, 0) == 1) |
| 426 |
\& continue; /* Retry */ |
| 427 |
\& printf("Failed to write end of HTTP request\en"); |
| 428 |
\& goto end; /* Cannot retry: error */ |
| 429 |
\& } |
| 430 |
.Ve |
| 431 |
.PP |
| 432 |
On a write we do not expect to see an EOF response so we treat that case in the |
| 433 |
same way as a fatal error. |
| 434 |
.PP |
| 435 |
Reading a response back from the server is similar: |
| 436 |
.PP |
| 437 |
.Vb 10 |
| 438 |
\& do { |
| 439 |
\& /* |
| 440 |
\& * Get up to sizeof(buf) bytes of the response. We keep reading until |
| 441 |
\& * the server closes the connection. |
| 442 |
\& */ |
| 443 |
\& while (!eof && !SSL_read_ex(ssl, buf, sizeof(buf), &readbytes)) { |
| 444 |
\& switch (handle_io_failure(ssl, 0)) { |
| 445 |
\& case 1: |
| 446 |
\& continue; /* Retry */ |
| 447 |
\& case 0: |
| 448 |
\& eof = 1; |
| 449 |
\& continue; |
| 450 |
\& case \-1: |
| 451 |
\& default: |
| 452 |
\& printf("Failed reading remaining data\en"); |
| 453 |
\& goto end; /* Cannot retry: error */ |
| 454 |
\& } |
| 455 |
\& } |
| 456 |
\& /* |
| 457 |
\& * OpenSSL does not guarantee that the returned data is a string or |
| 458 |
\& * that it is NUL terminated so we use fwrite() to write the exact |
| 459 |
\& * number of bytes that we read. The data could be non\-printable or |
| 460 |
\& * have NUL characters in the middle of it. For this simple example |
| 461 |
\& * we\*(Aqre going to print it to stdout anyway. |
| 462 |
\& */ |
| 463 |
\& if (!eof) |
| 464 |
\& fwrite(buf, 1, readbytes, stdout); |
| 465 |
\& } while (!eof); |
| 466 |
\& /* In case the response didn\*(Aqt finish with a newline we add one now */ |
| 467 |
\& printf("\en"); |
| 468 |
.Ve |
| 469 |
.PP |
| 470 |
The main difference this time is that it is valid for us to receive an EOF |
| 471 |
response when trying to read data from the server. This will occur when the |
| 472 |
server closes down the connection after sending all the data in its response. |
| 473 |
.PP |
| 474 |
In this demo we just print out all the data we've received back in the response |
| 475 |
from the server. We continue going around the loop until we either encounter a |
| 476 |
fatal error, or we receive an EOF (indicating a graceful finish). |
| 477 |
.SS "Shutting down the connection" |
| 478 |
.IX Subsection "Shutting down the connection" |
| 479 |
As in the QUIC blocking example we must shutdown the connection when we are |
| 480 |
finished with it. |
| 481 |
.PP |
| 482 |
Even though we have received EOF on the stream that we were reading from above, |
| 483 |
this tell us nothing about the state of the underlying connection. Our demo |
| 484 |
application will initiate the connection shutdown process via |
| 485 |
\&\fBSSL_shutdown\fR\|(3). |
| 486 |
.PP |
| 487 |
Since our application is initiating the shutdown then we might expect to see |
| 488 |
\&\fBSSL_shutdown\fR\|(3) give a return value of 0, and then we should continue to call |
| 489 |
it until we receive a return value of 1 (meaning we have successfully completed |
| 490 |
the shutdown). Since we are using a nonblocking \fBSSL\fR object we might expect to |
| 491 |
have to retry this operation several times. If \fBSSL_shutdown\fR\|(3) returns a |
| 492 |
negative result then we must call \fBSSL_get_error\fR\|(3) to work out what to do |
| 493 |
next. We use our \fBhandle_io_failure()\fR function that we developed earlier for |
| 494 |
this: |
| 495 |
.PP |
| 496 |
.Vb 8 |
| 497 |
\& /* |
| 498 |
\& * Repeatedly call SSL_shutdown() until the connection is fully |
| 499 |
\& * closed. |
| 500 |
\& */ |
| 501 |
\& while ((ret = SSL_shutdown(ssl)) != 1) { |
| 502 |
\& if (ret < 0 && handle_io_failure(ssl, ret) == 1) |
| 503 |
\& continue; /* Retry */ |
| 504 |
\& } |
| 505 |
.Ve |
| 506 |
.SS "Final clean up" |
| 507 |
.IX Subsection "Final clean up" |
| 508 |
As with the blocking QUIC client example, once our connection is finished with |
| 509 |
we must free it. The steps to do this for this example are the same as for the |
| 510 |
blocking example, so we won't repeat it here. |
| 511 |
.SH "FURTHER READING" |
| 512 |
.IX Header "FURTHER READING" |
| 513 |
See \fBossl\-guide\-quic\-client\-block\fR\|(7) to read a tutorial on how to write a |
| 514 |
blocking QUIC client. See \fBossl\-guide\-quic\-multi\-stream\fR\|(7) to see how to write |
| 515 |
a multi-stream QUIC client. |
| 516 |
.SH "SEE ALSO" |
| 517 |
.IX Header "SEE ALSO" |
| 518 |
\&\fBossl\-guide\-introduction\fR\|(7), \fBossl\-guide\-libraries\-introduction\fR\|(7), |
| 519 |
\&\fBossl\-guide\-libssl\-introduction\fR\|(7), \fBossl\-guide\-quic\-introduction\fR\|(7), |
| 520 |
\&\fBossl\-guide\-quic\-client\-block\fR\|(7), \fBossl\-guide\-quic\-multi\-stream\fR\|(7) |
| 521 |
.SH COPYRIGHT |
| 522 |
.IX Header "COPYRIGHT" |
| 523 |
Copyright 2023\-2025 The OpenSSL Project Authors. All Rights Reserved. |
| 524 |
.PP |
| 525 |
Licensed under the Apache License 2.0 (the "License"). You may not use |
| 526 |
this file except in compliance with the License. You can obtain a copy |
| 527 |
in the file LICENSE in the source distribution or at |
| 528 |
<https://www.openssl.org/source/license.html>. |