Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 1 | =pod |
| 2 | |
| 3 | =head1 NAME |
| 4 | |
Richard Levitte | 8eec138 | 2000-09-14 20:24:56 +0000 | [diff] [blame] | 5 | BIO_s_bio, BIO_make_bio_pair, BIO_destroy_bio_pair, BIO_set_write_buf_size, |
| 6 | BIO_get_write_buf_size, BIO_new_bio_pair, BIO_get_write_guarantee, |
| 7 | BIO_ctrl_get_write_guarantee, BIO_get_read_request, BIO_ctrl_get_read_request, |
| 8 | BIO_ctrl_reset_read_request - BIO pair BIO |
Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 9 | |
| 10 | =head1 SYNOPSIS |
| 11 | |
| 12 | #include <openssl/bio.h> |
| 13 | |
| 14 | BIO_METHOD *BIO_s_bio(void); |
| 15 | |
| 16 | #define BIO_make_bio_pair(b1,b2) (int)BIO_ctrl(b1,BIO_C_MAKE_BIO_PAIR,0,b2) |
| 17 | #define BIO_destroy_bio_pair(b) (int)BIO_ctrl(b,BIO_C_DESTROY_BIO_PAIR,0,NULL) |
| 18 | |
| 19 | #define BIO_set_write_buf_size(b,size) (int)BIO_ctrl(b,BIO_C_SET_WRITE_BUF_SIZE,size,NULL) |
| 20 | #define BIO_get_write_buf_size(b,size) (size_t)BIO_ctrl(b,BIO_C_GET_WRITE_BUF_SIZE,size,NULL) |
| 21 | |
| 22 | int BIO_new_bio_pair(BIO **bio1, size_t writebuf1, BIO **bio2, size_t writebuf2); |
| 23 | |
| 24 | #define BIO_get_write_guarantee(b) (int)BIO_ctrl(b,BIO_C_GET_WRITE_GUARANTEE,0,NULL) |
| 25 | size_t BIO_ctrl_get_write_guarantee(BIO *b); |
| 26 | |
| 27 | #define BIO_get_read_request(b) (int)BIO_ctrl(b,BIO_C_GET_READ_REQUEST,0,NULL) |
| 28 | size_t BIO_ctrl_get_read_request(BIO *b); |
| 29 | |
| 30 | int BIO_ctrl_reset_read_request(BIO *b); |
| 31 | |
| 32 | =head1 DESCRIPTION |
| 33 | |
| 34 | BIO_s_bio() returns the method for a BIO pair. A BIO pair is a pair of source/sink |
| 35 | BIOs where data written to either half of the pair is buffered and can be read from |
Bodo Möller | e39c194 | 2000-09-14 22:09:55 +0000 | [diff] [blame^] | 36 | the other half. Both halves must usually by handled by the same application thread |
| 37 | since no locking is done on the internal data structures. |
Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 38 | |
| 39 | Since BIO chains typically end in a source/sink BIO it is possible to make this |
| 40 | one half of a BIO pair and have all the data processed by the chain under application |
| 41 | control. |
| 42 | |
Bodo Möller | e39c194 | 2000-09-14 22:09:55 +0000 | [diff] [blame^] | 43 | One typical use of BIO pairs is to place TLS/SSL I/O under application control, this |
| 44 | can be used when the application wishes to use a non standard transport for |
| 45 | TLS/SSL or the normal socket routines are inappropriate. |
Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 46 | |
| 47 | Calls to BIO_read() will read data from the buffer or request a retry if no |
| 48 | data is available. |
| 49 | |
| 50 | Calls to BIO_write() will place data in the buffer or request a retry if the |
| 51 | buffer is full. |
| 52 | |
| 53 | The standard calls BIO_ctrl_pending() and BIO_ctrl_wpending() can be used to |
| 54 | determine the amount of pending data in the read or write buffer. |
| 55 | |
| 56 | BIO_reset() clears any data in the write buffer. |
| 57 | |
| 58 | BIO_make_bio_pair() joins two separate BIOs into a connected pair. |
| 59 | |
| 60 | BIO_destroy_pair() destroys the association between two connected BIOs. Freeing |
Bodo Möller | e39c194 | 2000-09-14 22:09:55 +0000 | [diff] [blame^] | 61 | up any half of the pair will automatically destroy the association. |
Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 62 | |
| 63 | BIO_set_write_buf_size() sets the write buffer size of BIO B<b> to B<size>. |
| 64 | If the size is not initialised a default value is used. This is currently |
| 65 | 17K, sufficient for a maximum size TLS record. |
| 66 | |
| 67 | BIO_get_write_buf_size() returns the size of the write buffer. |
| 68 | |
| 69 | BIO_new_bio_pair() combines the calls to BIO_new(), BIO_make_bio_pair() and |
| 70 | BIO_set_write_buf_size() to create a connected pair of BIOs B<bio1>, B<bio2> |
| 71 | with write buffer sizes B<writebuf1> and B<writebuf2>. If either size is |
| 72 | zero then the default size is used. |
| 73 | |
| 74 | BIO_get_write_guarantee() and BIO_ctrl_get_write_guarentee() return the maximum |
| 75 | length of data that can be currently written to the BIO. Writes larger than this |
| 76 | value will return a value from BIO_write() less than the amount requested or if the |
| 77 | buffer is full request a retry. BIO_ctrl_get_write_guarantee() is a function |
| 78 | whereas BIO_get_write_guarantee() is a macro. |
| 79 | |
| 80 | BIO_get_read_request() and BIO_ctrl_get_read_request() return the amount of data |
| 81 | requested (or the buffer size if it is less) if the last read failed due to an |
| 82 | empty buffer. This can be used to determine how much data should be written to the |
Bodo Möller | e39c194 | 2000-09-14 22:09:55 +0000 | [diff] [blame^] | 83 | other half of the pair so the next read will succeed: this is most useful in TLS/SSL |
Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 84 | applications where the amount of data read is usually meaningful rather than just |
| 85 | a buffer size. After a successful read this call will return zero. |
| 86 | |
| 87 | BIO_ctrl_reset_read_request() can also be used to reset the value returned by |
| 88 | BIO_get_read_request() to zero. |
| 89 | |
| 90 | =head1 NOTES |
| 91 | |
| 92 | Both halves of a BIO pair should be freed. That is even if one half is implicity |
| 93 | freed due to a BIO_free_all() or SSL_free() call the other half needs to be freed. |
| 94 | |
Bodo Möller | e39c194 | 2000-09-14 22:09:55 +0000 | [diff] [blame^] | 95 | When used in bidirectional applications (such as TLS/SSL) care should be taken to |
Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 96 | flush any data in the write buffer. This can be done by calling BIO_pending() |
| 97 | on the other half of the pair and, if any data is pending, reading it and sending |
| 98 | it to the underlying transport. This must be done before any normal processing |
| 99 | (such as calling select() ) due to a request and BIO_should_read() being true. |
| 100 | |
| 101 | To see why this is important consider a case where a request is sent using |
| 102 | BIO_write() and a response read with BIO_read(), this can occur during an |
Bodo Möller | e39c194 | 2000-09-14 22:09:55 +0000 | [diff] [blame^] | 103 | TLS/SSL handshake for example. BIO_write() will succeed and place data in the write |
Dr. Stephen Henson | 18f2259 | 2000-09-14 18:55:39 +0000 | [diff] [blame] | 104 | buffer. BIO_read() will initially fail and BIO_should_read() will be true. If |
| 105 | the application then waits for data to be available on the underlying transport |
| 106 | before flusing the write buffer it will never succeed because the request was |
| 107 | never sent! |
| 108 | |
| 109 | =head1 EXAMPLE |
| 110 | |
| 111 | TBA |
| 112 | |
| 113 | =head1 SEE ALSO |
| 114 | |
| 115 | L<SSL_set_bio(3)|SSL_set_bio(3)>, L<ssl(3)|ssl(3)>, L<bio(3)|bio(3)>, |
| 116 | L<BIO_should_retry(3)|BIO_should_retry(3)>, L<BIO_read(3)|BIO_read(3)> |
| 117 | |
| 118 | =cut |