1*2175Sjp161948=pod 2*2175Sjp161948 3*2175Sjp161948=head1 NAME 4*2175Sjp161948 5*2175Sjp161948BIO_s_bio, BIO_make_bio_pair, BIO_destroy_bio_pair, BIO_shutdown_wr, 6*2175Sjp161948BIO_set_write_buf_size, BIO_get_write_buf_size, BIO_new_bio_pair, 7*2175Sjp161948BIO_get_write_guarantee, BIO_ctrl_get_write_guarantee, BIO_get_read_request, 8*2175Sjp161948BIO_ctrl_get_read_request, BIO_ctrl_reset_read_request - BIO pair BIO 9*2175Sjp161948 10*2175Sjp161948=head1 SYNOPSIS 11*2175Sjp161948 12*2175Sjp161948 #include <openssl/bio.h> 13*2175Sjp161948 14*2175Sjp161948 BIO_METHOD *BIO_s_bio(void); 15*2175Sjp161948 16*2175Sjp161948 #define BIO_make_bio_pair(b1,b2) (int)BIO_ctrl(b1,BIO_C_MAKE_BIO_PAIR,0,b2) 17*2175Sjp161948 #define BIO_destroy_bio_pair(b) (int)BIO_ctrl(b,BIO_C_DESTROY_BIO_PAIR,0,NULL) 18*2175Sjp161948 19*2175Sjp161948 #define BIO_shutdown_wr(b) (int)BIO_ctrl(b, BIO_C_SHUTDOWN_WR, 0, NULL) 20*2175Sjp161948 21*2175Sjp161948 #define BIO_set_write_buf_size(b,size) (int)BIO_ctrl(b,BIO_C_SET_WRITE_BUF_SIZE,size,NULL) 22*2175Sjp161948 #define BIO_get_write_buf_size(b,size) (size_t)BIO_ctrl(b,BIO_C_GET_WRITE_BUF_SIZE,size,NULL) 23*2175Sjp161948 24*2175Sjp161948 int BIO_new_bio_pair(BIO **bio1, size_t writebuf1, BIO **bio2, size_t writebuf2); 25*2175Sjp161948 26*2175Sjp161948 #define BIO_get_write_guarantee(b) (int)BIO_ctrl(b,BIO_C_GET_WRITE_GUARANTEE,0,NULL) 27*2175Sjp161948 size_t BIO_ctrl_get_write_guarantee(BIO *b); 28*2175Sjp161948 29*2175Sjp161948 #define BIO_get_read_request(b) (int)BIO_ctrl(b,BIO_C_GET_READ_REQUEST,0,NULL) 30*2175Sjp161948 size_t BIO_ctrl_get_read_request(BIO *b); 31*2175Sjp161948 32*2175Sjp161948 int BIO_ctrl_reset_read_request(BIO *b); 33*2175Sjp161948 34*2175Sjp161948=head1 DESCRIPTION 35*2175Sjp161948 36*2175Sjp161948BIO_s_bio() returns the method for a BIO pair. A BIO pair is a pair of source/sink 37*2175Sjp161948BIOs where data written to either half of the pair is buffered and can be read from 38*2175Sjp161948the other half. Both halves must usually by handled by the same application thread 39*2175Sjp161948since no locking is done on the internal data structures. 40*2175Sjp161948 41*2175Sjp161948Since BIO chains typically end in a source/sink BIO it is possible to make this 42*2175Sjp161948one half of a BIO pair and have all the data processed by the chain under application 43*2175Sjp161948control. 44*2175Sjp161948 45*2175Sjp161948One typical use of BIO pairs is to place TLS/SSL I/O under application control, this 46*2175Sjp161948can be used when the application wishes to use a non standard transport for 47*2175Sjp161948TLS/SSL or the normal socket routines are inappropriate. 48*2175Sjp161948 49*2175Sjp161948Calls to BIO_read() will read data from the buffer or request a retry if no 50*2175Sjp161948data is available. 51*2175Sjp161948 52*2175Sjp161948Calls to BIO_write() will place data in the buffer or request a retry if the 53*2175Sjp161948buffer is full. 54*2175Sjp161948 55*2175Sjp161948The standard calls BIO_ctrl_pending() and BIO_ctrl_wpending() can be used to 56*2175Sjp161948determine the amount of pending data in the read or write buffer. 57*2175Sjp161948 58*2175Sjp161948BIO_reset() clears any data in the write buffer. 59*2175Sjp161948 60*2175Sjp161948BIO_make_bio_pair() joins two separate BIOs into a connected pair. 61*2175Sjp161948 62*2175Sjp161948BIO_destroy_pair() destroys the association between two connected BIOs. Freeing 63*2175Sjp161948up any half of the pair will automatically destroy the association. 64*2175Sjp161948 65*2175Sjp161948BIO_shutdown_wr() is used to close down a BIO B<b>. After this call no further 66*2175Sjp161948writes on BIO B<b> are allowed (they will return an error). Reads on the other 67*2175Sjp161948half of the pair will return any pending data or EOF when all pending data has 68*2175Sjp161948been read. 69*2175Sjp161948 70*2175Sjp161948BIO_set_write_buf_size() sets the write buffer size of BIO B<b> to B<size>. 71*2175Sjp161948If the size is not initialized a default value is used. This is currently 72*2175Sjp16194817K, sufficient for a maximum size TLS record. 73*2175Sjp161948 74*2175Sjp161948BIO_get_write_buf_size() returns the size of the write buffer. 75*2175Sjp161948 76*2175Sjp161948BIO_new_bio_pair() combines the calls to BIO_new(), BIO_make_bio_pair() and 77*2175Sjp161948BIO_set_write_buf_size() to create a connected pair of BIOs B<bio1>, B<bio2> 78*2175Sjp161948with write buffer sizes B<writebuf1> and B<writebuf2>. If either size is 79*2175Sjp161948zero then the default size is used. BIO_new_bio_pair() does not check whether 80*2175Sjp161948B<bio1> or B<bio2> do point to some other BIO, the values are overwritten, 81*2175Sjp161948BIO_free() is not called. 82*2175Sjp161948 83*2175Sjp161948BIO_get_write_guarantee() and BIO_ctrl_get_write_guarantee() return the maximum 84*2175Sjp161948length of data that can be currently written to the BIO. Writes larger than this 85*2175Sjp161948value will return a value from BIO_write() less than the amount requested or if the 86*2175Sjp161948buffer is full request a retry. BIO_ctrl_get_write_guarantee() is a function 87*2175Sjp161948whereas BIO_get_write_guarantee() is a macro. 88*2175Sjp161948 89*2175Sjp161948BIO_get_read_request() and BIO_ctrl_get_read_request() return the 90*2175Sjp161948amount of data requested, or the buffer size if it is less, if the 91*2175Sjp161948last read attempt at the other half of the BIO pair failed due to an 92*2175Sjp161948empty buffer. This can be used to determine how much data should be 93*2175Sjp161948written to the BIO so the next read will succeed: this is most useful 94*2175Sjp161948in TLS/SSL applications where the amount of data read is usually 95*2175Sjp161948meaningful rather than just a buffer size. After a successful read 96*2175Sjp161948this call will return zero. It also will return zero once new data 97*2175Sjp161948has been written satisfying the read request or part of it. 98*2175Sjp161948Note that BIO_get_read_request() never returns an amount larger 99*2175Sjp161948than that returned by BIO_get_write_guarantee(). 100*2175Sjp161948 101*2175Sjp161948BIO_ctrl_reset_read_request() can also be used to reset the value returned by 102*2175Sjp161948BIO_get_read_request() to zero. 103*2175Sjp161948 104*2175Sjp161948=head1 NOTES 105*2175Sjp161948 106*2175Sjp161948Both halves of a BIO pair should be freed. That is even if one half is implicit 107*2175Sjp161948freed due to a BIO_free_all() or SSL_free() call the other half needs to be freed. 108*2175Sjp161948 109*2175Sjp161948When used in bidirectional applications (such as TLS/SSL) care should be taken to 110*2175Sjp161948flush any data in the write buffer. This can be done by calling BIO_pending() 111*2175Sjp161948on the other half of the pair and, if any data is pending, reading it and sending 112*2175Sjp161948it to the underlying transport. This must be done before any normal processing 113*2175Sjp161948(such as calling select() ) due to a request and BIO_should_read() being true. 114*2175Sjp161948 115*2175Sjp161948To see why this is important consider a case where a request is sent using 116*2175Sjp161948BIO_write() and a response read with BIO_read(), this can occur during an 117*2175Sjp161948TLS/SSL handshake for example. BIO_write() will succeed and place data in the write 118*2175Sjp161948buffer. BIO_read() will initially fail and BIO_should_read() will be true. If 119*2175Sjp161948the application then waits for data to be available on the underlying transport 120*2175Sjp161948before flushing the write buffer it will never succeed because the request was 121*2175Sjp161948never sent! 122*2175Sjp161948 123*2175Sjp161948=head1 RETURN VALUES 124*2175Sjp161948 125*2175Sjp161948BIO_new_bio_pair() returns 1 on success, with the new BIOs available in 126*2175Sjp161948B<bio1> and B<bio2>, or 0 on failure, with NULL pointers stored into the 127*2175Sjp161948locations for B<bio1> and B<bio2>. Check the error stack for more information. 128*2175Sjp161948 129*2175Sjp161948[XXXXX: More return values need to be added here] 130*2175Sjp161948 131*2175Sjp161948=head1 EXAMPLE 132*2175Sjp161948 133*2175Sjp161948The BIO pair can be used to have full control over the network access of an 134*2175Sjp161948application. The application can call select() on the socket as required 135*2175Sjp161948without having to go through the SSL-interface. 136*2175Sjp161948 137*2175Sjp161948 BIO *internal_bio, *network_bio; 138*2175Sjp161948 ... 139*2175Sjp161948 BIO_new_bio_pair(internal_bio, 0, network_bio, 0); 140*2175Sjp161948 SSL_set_bio(ssl, internal_bio, internal_bio); 141*2175Sjp161948 SSL_operations(); 142*2175Sjp161948 ... 143*2175Sjp161948 144*2175Sjp161948 application | TLS-engine 145*2175Sjp161948 | | 146*2175Sjp161948 +----------> SSL_operations() 147*2175Sjp161948 | /\ || 148*2175Sjp161948 | || \/ 149*2175Sjp161948 | BIO-pair (internal_bio) 150*2175Sjp161948 +----------< BIO-pair (network_bio) 151*2175Sjp161948 | | 152*2175Sjp161948 socket | 153*2175Sjp161948 154*2175Sjp161948 ... 155*2175Sjp161948 SSL_free(ssl); /* implicitly frees internal_bio */ 156*2175Sjp161948 BIO_free(network_bio); 157*2175Sjp161948 ... 158*2175Sjp161948 159*2175Sjp161948As the BIO pair will only buffer the data and never directly access the 160*2175Sjp161948connection, it behaves non-blocking and will return as soon as the write 161*2175Sjp161948buffer is full or the read buffer is drained. Then the application has to 162*2175Sjp161948flush the write buffer and/or fill the read buffer. 163*2175Sjp161948 164*2175Sjp161948Use the BIO_ctrl_pending(), to find out whether data is buffered in the BIO 165*2175Sjp161948and must be transfered to the network. Use BIO_ctrl_get_read_request() to 166*2175Sjp161948find out, how many bytes must be written into the buffer before the 167*2175Sjp161948SSL_operation() can successfully be continued. 168*2175Sjp161948 169*2175Sjp161948=head1 WARNING 170*2175Sjp161948 171*2175Sjp161948As the data is buffered, SSL_operation() may return with a ERROR_SSL_WANT_READ 172*2175Sjp161948condition, but there is still data in the write buffer. An application must 173*2175Sjp161948not rely on the error value of SSL_operation() but must assure that the 174*2175Sjp161948write buffer is always flushed first. Otherwise a deadlock may occur as 175*2175Sjp161948the peer might be waiting for the data before being able to continue. 176*2175Sjp161948 177*2175Sjp161948=head1 SEE ALSO 178*2175Sjp161948 179*2175Sjp161948L<SSL_set_bio(3)|SSL_set_bio(3)>, L<ssl(3)|ssl(3)>, L<bio(3)|bio(3)>, 180*2175Sjp161948L<BIO_should_retry(3)|BIO_should_retry(3)>, L<BIO_read(3)|BIO_read(3)> 181*2175Sjp161948 182*2175Sjp161948=cut 183