Quelle ssl_engine_io.c Sprache: C

/* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements.  See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License.  You may obtain a copy of the License at
*
*     http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/

/*                      _             _
*  _ __ ___   ___   __| |    ___ ___| |  mod_ssl
* | '_ ` _ \ / _ \ / _` |   / __/ __| |  Apache Interface to OpenSSL
* | | | | | | (_) | (_| |   \__ \__ \ |
* |_| |_| |_|\___/ \__,_|___|___/___/_|
*                      |_____|
*  ssl_engine_io.c
*  I/O Functions
*/
                             /* ``MY HACK: This universe.
                                  Just one little problem:
                                  core keeps dumping.''
                                            -- Unknown    */
#include "ssl_private.h"

#include "apr_date.h"

APR_IMPLEMENT_OPTIONAL_HOOK_RUN_ALL(ssl, SSL, int, proxy_post_handshake,
                                    (conn_rec *c,SSL *ssl),
                                    (c,ssl),OK,DECLINED);

/*  _________________________________________________________________
**
**  I/O Hooks
**  _________________________________________________________________
*/

/* This file is designed to be the bridge between OpenSSL and httpd.
* However, we really don't expect anyone (let alone ourselves) to
* remember what is in this file.  So, first, a quick overview.
*
* In this file, you will find:
* - ssl_io_filter_input    (Apache input filter)
* - ssl_io_filter_output   (Apache output filter)
*
* - bio_filter_in_*        (OpenSSL input filter)
* - bio_filter_out_*       (OpenSSL output filter)
*
* The input chain is roughly:
*
* ssl_io_filter_input->ssl_io_input_read->SSL_read->...
* ...->bio_filter_in_read->ap_get_brigade/next-httpd-filter
*
* In mortal terminology, we do the following:
* - Receive a request for data to the SSL input filter
* - Call a helper function once we know we should perform a read
* - Call OpenSSL's SSL_read()
* - SSL_read() will then call bio_filter_in_read
* - bio_filter_in_read will then try to fetch data from the next httpd filter
* - bio_filter_in_read will flatten that data and return it to SSL_read
* - SSL_read will then decrypt the data
* - ssl_io_input_read will then receive decrypted data as a char* and
*   ensure that there were no read errors
* - The char* is placed in a brigade and returned
*
* Since connection-level input filters in httpd need to be able to
* handle AP_MODE_GETLINE calls (namely identifying LF-terminated strings),
* ssl_io_input_getline which will handle this special case.
*
* Due to AP_MODE_GETLINE and AP_MODE_SPECULATIVE, we may sometimes have
* 'leftover' decoded data which must be setaside for the next read.  That
* is currently handled by the char_buffer_{read|write} functions.  So,
* ssl_io_input_read may be able to fulfill reads without invoking
* SSL_read().
*
* Note that the filter context of ssl_io_filter_input and bio_filter_in_*
* are shared as bio_filter_in_ctx_t.
*
* Note that the filter is by choice limited to reading at most
* AP_IOBUFSIZE (8192 bytes) per call.
*
*/

/* this custom BIO allows us to hook SSL_write directly into
* an apr_bucket_brigade and use transient buckets with the SSL
* malloc-ed buffer, rather than copying into a mem BIO.
* also allows us to pass the brigade as data is being written
* rather than buffering up the entire response in the mem BIO.
*
* when SSL needs to flush (e.g. SSL_accept()), it will call BIO_flush()
* which will trigger a call to bio_filter_out_ctrl() -> bio_filter_out_flush().
* so we only need to flush the output ourselves if we receive an
* EOS or FLUSH bucket. this was not possible with the mem BIO where we
* had to flush all over the place not really knowing when it was required
* to do so.
*/

typedef struct {
    SSL                *pssl;
    BIO                *pbioRead;
    BIO                *pbioWrite;
    ap_filter_t        *pInputFilter;
    ap_filter_t        *pOutputFilter;
    SSLConnRec         *config;
} ssl_filter_ctx_t;

typedef struct {
    ssl_filter_ctx_t *filter_ctx;
    conn_rec *c;
    apr_bucket_brigade *bb;    /* Brigade used as a buffer. */
    apr_status_t rc;
} bio_filter_out_ctx_t;

static bio_filter_out_ctx_t *bio_filter_out_ctx_new(ssl_filter_ctx_t *filter_ctx,
                                                    conn_rec *c)
{
    bio_filter_out_ctx_t *outctx = apr_palloc(c->pool, sizeof(*outctx));

    outctx->filter_ctx = filter_ctx;
    outctx->c = c;
    outctx->bb = apr_brigade_create(c->pool, c->bucket_alloc);

    return outctx;
}

/* Pass an output brigade down the filter stack; returns 1 on success
* or -1 on failure. */
static int bio_filter_out_pass(bio_filter_out_ctx_t *outctx)
{
    AP_DEBUG_ASSERT(!APR_BRIGADE_EMPTY(outctx->bb));

    outctx->rc = ap_pass_brigade(outctx->filter_ctx->pOutputFilter->next,
                                 outctx->bb);
    /* Fail if the connection was reset: */
    if (outctx->rc == APR_SUCCESS && outctx->c->aborted) {
        outctx->rc = APR_ECONNRESET;
    }
    return (outctx->rc == APR_SUCCESS) ? 1 : -1;
}

/* Send a FLUSH bucket down the output filter stack; returns 1 on
* success, -1 on failure. */
static int bio_filter_out_flush(BIO *bio)
{
    bio_filter_out_ctx_t *outctx = (bio_filter_out_ctx_t *)BIO_get_data(bio);
    apr_bucket *e;

    ap_log_cerror(APLOG_MARK, APLOG_TRACE6, 0, outctx->c,
                  "bio_filter_out_write: flush");

    AP_DEBUG_ASSERT(APR_BRIGADE_EMPTY(outctx->bb));

    e = apr_bucket_flush_create(outctx->bb->bucket_alloc);
    APR_BRIGADE_INSERT_TAIL(outctx->bb, e);

    return bio_filter_out_pass(outctx);
}

static int bio_filter_create(BIO *bio)
{
    BIO_set_shutdown(bio, 1);
    BIO_set_init(bio, 1);
#if MODSSL_USE_OPENSSL_PRE_1_1_API
    /* No setter method for OpenSSL 1.1.0 available,
     * but I can't find any functional use of the
     * "num" field there either.
     */
    bio->num = -1;
#endif
    BIO_set_data(bio, NULL);

    return 1;
}

static int bio_filter_destroy(BIO *bio)
{
    if (bio == NULL) {
        return 0;
    }

    /* nothing to free here.
     * apache will destroy the bucket brigade for us
     */
    return 1;
}

static int bio_filter_out_read(BIO *bio, char *out, int outl)
{
    /* this is never called */
    bio_filter_out_ctx_t *outctx = (bio_filter_out_ctx_t *)BIO_get_data(bio);
    ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, outctx->c,
                  "BUG: %s() should not be called", "bio_filter_out_read");
    AP_DEBUG_ASSERT(0);
    return -1;
}

static int bio_filter_out_write(BIO *bio, const char *in, int inl)
{
    bio_filter_out_ctx_t *outctx = (bio_filter_out_ctx_t *)BIO_get_data(bio);
    apr_bucket *e;
    int need_flush;

    BIO_clear_retry_flags(bio);

#ifndef SSL_OP_NO_RENEGOTIATION
    /* Abort early if the client has initiated a renegotiation. */
    if (outctx->filter_ctx->config->reneg_state == RENEG_ABORT) {
        outctx->rc = APR_ECONNABORTED;
        return -1;
    }
#endif

    ap_log_cerror(APLOG_MARK, APLOG_TRACE6, 0, outctx->c,
                  "bio_filter_out_write: %i bytes", inl);

    /* Use a transient bucket for the output data - any downstream
     * filter must setaside if necessary. */
    e = apr_bucket_transient_create(in, inl, outctx->bb->bucket_alloc);
    APR_BRIGADE_INSERT_TAIL(outctx->bb, e);

    /* In theory, OpenSSL should flush as necessary, but it is known
     * not to do so correctly in some cases (< 0.9.8m; see PR 46952),
     * or on the proxy/client side (after ssl23_client_hello(), e.g.
     * ssl/proxy.t test suite).
     *
     * Historically, this flush call was performed only for an SSLv2
     * connection or for a proxy connection.  Calling _out_flush can
     * be expensive in cases where requests/responses are pipelined,
     * so limit the performance impact to handshake time.
     */
#if OPENSSL_VERSION_NUMBER < 0x0009080df
     need_flush = !SSL_is_init_finished(outctx->filter_ctx->pssl);
#else
     need_flush = SSL_in_connect_init(outctx->filter_ctx->pssl);
#endif
    if (need_flush) {
        e = apr_bucket_flush_create(outctx->bb->bucket_alloc);
        APR_BRIGADE_INSERT_TAIL(outctx->bb, e);
    }

    if (bio_filter_out_pass(outctx) < 0) {
        return -1;
    }

    return inl;
}

static long bio_filter_out_ctrl(BIO *bio, int cmd, long num, void *ptr)
{
    long ret = 1;
    bio_filter_out_ctx_t *outctx = (bio_filter_out_ctx_t *)BIO_get_data(bio);

    switch (cmd) {
    case BIO_CTRL_RESET:
    case BIO_CTRL_EOF:
    case BIO_C_SET_BUF_MEM_EOF_RETURN:
        ap_log_cerror(APLOG_MARK, APLOG_TRACE4, 0, outctx->c,
                      "output bio: unhandled control %d", cmd);
        ret = 0;
        break;
    case BIO_CTRL_WPENDING:
    case BIO_CTRL_PENDING:
    case BIO_CTRL_INFO:
        ret = 0;
        break;
    case BIO_CTRL_GET_CLOSE:
        ret = (long)BIO_get_shutdown(bio);
        break;
      case BIO_CTRL_SET_CLOSE:
        BIO_set_shutdown(bio, (int)num);
        break;
      case BIO_CTRL_FLUSH:
        ret = bio_filter_out_flush(bio);
        break;
      case BIO_CTRL_DUP:
        ret = 1;
        break;
        /* N/A */
      case BIO_C_SET_BUF_MEM:
      case BIO_C_GET_BUF_MEM_PTR:
        /* we don't care */
      case BIO_CTRL_PUSH:
      case BIO_CTRL_POP:
      default:
        ret = 0;
        break;
    }

    return ret;
}

static int bio_filter_out_gets(BIO *bio, char *buf, int size)
{
    /* this is never called */
    bio_filter_out_ctx_t *outctx = (bio_filter_out_ctx_t *)BIO_get_data(bio);
    ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, outctx->c,
                  "BUG: %s() should not be called", "bio_filter_out_gets");
    AP_DEBUG_ASSERT(0);
    return -1;
}

static int bio_filter_out_puts(BIO *bio, const char *str)
{
    /* this is never called */
    bio_filter_out_ctx_t *outctx = (bio_filter_out_ctx_t *)BIO_get_data(bio);
    ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, outctx->c,
                  "BUG: %s() should not be called", "bio_filter_out_puts");
    AP_DEBUG_ASSERT(0);
    return -1;
}

typedef struct {
    int length;
    char *value;
} char_buffer_t;

typedef struct {
    SSL *ssl;
    BIO *bio_out;
    ap_filter_t *f;
    apr_status_t rc;
    ap_input_mode_t mode;
    apr_read_type_e block;
    apr_bucket_brigade *bb;
    char_buffer_t cbuf;
    apr_pool_t *pool;
    char buffer[AP_IOBUFSIZE];
    ssl_filter_ctx_t *filter_ctx;
} bio_filter_in_ctx_t;

/*
* this char_buffer api might seem silly, but we don't need to copy
* any of this data and we need to remember the length.
*/

/* Copy up to INL bytes from the char_buffer BUFFER into IN.  Note
* that due to the strange way this API is designed/used, the
* char_buffer object is used to cache a segment of inctx->buffer, and
* then this function called to copy (part of) that segment to the
* beginning of inctx->buffer.  So the segments to copy cannot be
* presumed to be non-overlapping, and memmove must be used. */
static int char_buffer_read(char_buffer_t *buffer, char *in, int inl)
{
    if (!buffer->length) {
        return 0;
    }

    if (buffer->length > inl) {
        /* we have enough to fill the caller's buffer */
        memmove(in, buffer->value, inl);
        buffer->value += inl;
        buffer->length -= inl;
    }
    else {
        /* swallow remainder of the buffer */
        memmove(in, buffer->value, buffer->length);
        inl = buffer->length;
        buffer->value = NULL;
        buffer->length = 0;
    }

    return inl;
}

static int char_buffer_write(char_buffer_t *buffer, char *in, int inl)
{
    buffer->value = in;
    buffer->length = inl;
    return inl;
}

/* This function will read from a brigade and discard the read buckets as it
* proceeds.  It will read at most *len bytes.
*/
static apr_status_t brigade_consume(apr_bucket_brigade *bb,
                                    apr_read_type_e block,
                                    char *c, apr_size_t *len)
{
    apr_size_t actual = 0;
    apr_status_t status = APR_SUCCESS;

    while (!APR_BRIGADE_EMPTY(bb)) {
        apr_bucket *b = APR_BRIGADE_FIRST(bb);
        const char *str;
        apr_size_t str_len;
        apr_size_t consume;

        /* Justin points out this is an http-ism that might
         * not fit if brigade_consume is added to APR.  Perhaps
         * apr_bucket_read(eos_bucket) should return APR_EOF?
         * Then this becomes mainline instead of a one-off.
         */
        if (APR_BUCKET_IS_EOS(b)) {
            status = APR_EOF;
            break;
        }

        /* The reason I'm not offering brigade_consume yet
         * across to apr-util is that the following call
         * illustrates how borked that API really is.  For
         * this sort of case (caller provided buffer) it
         * would be much more trivial for apr_bucket_consume
         * to do all the work that follows, based on the
         * particular characteristics of the bucket we are
         * consuming here.
         */
        status = apr_bucket_read(b, &str, &str_len, block);

        if (status != APR_SUCCESS) {
            if (APR_STATUS_IS_EOF(status)) {
                /* This stream bucket was consumed */
                apr_bucket_delete(b);
                continue;
            }
            break;
        }

        if (str_len > 0) {
            /* Do not block once some data has been consumed */
            block = APR_NONBLOCK_READ;

            /* Assure we don't overflow. */
            consume = (str_len + actual > *len) ? *len - actual : str_len;

            memcpy(c, str, consume);

            c += consume;
            actual += consume;

            if (consume >= b->length) {
                /* This physical bucket was consumed */
                apr_bucket_delete(b);
            }
            else {
                /* Only part of this physical bucket was consumed */
                b->start += consume;
                b->length -= consume;
            }
        }
        else if (b->length == 0) {
            apr_bucket_delete(b);
        }

        /* This could probably be actual == *len, but be safe from stray
         * photons. */
        if (actual >= *len) {
            break;
        }
    }

    *len = actual;
    return status;
}

/*
* this is the function called by SSL_read()
*/
static int bio_filter_in_read(BIO *bio, char *in, int inlen)
{
    apr_size_t inl = inlen;
    bio_filter_in_ctx_t *inctx = (bio_filter_in_ctx_t *)BIO_get_data(bio);
    apr_read_type_e block = inctx->block;

    inctx->rc = APR_SUCCESS;

    /* OpenSSL catches this case, so should we. */
    if (!in)
        return 0;

    BIO_clear_retry_flags(bio);

#ifndef SSL_OP_NO_RENEGOTIATION
    /* Abort early if the client has initiated a renegotiation. */
    if (inctx->filter_ctx->config->reneg_state == RENEG_ABORT) {
        inctx->rc = APR_ECONNABORTED;
        return -1;
    }
#endif

    if (!inctx->bb) {
        inctx->rc = APR_EOF;
        return -1;
    }

    if (APR_BRIGADE_EMPTY(inctx->bb)) {

        inctx->rc = ap_get_brigade(inctx->f->next, inctx->bb,
                                   AP_MODE_READBYTES, block,
                                   inl);

        /* If the read returns EAGAIN or success with an empty
         * brigade, return an error after setting the retry flag;
         * SSL_read() will then return -1, and SSL_get_error() will
         * indicate SSL_ERROR_WANT_READ. */
        if (APR_STATUS_IS_EAGAIN(inctx->rc) || APR_STATUS_IS_EINTR(inctx->rc)
               || (inctx->rc == APR_SUCCESS && APR_BRIGADE_EMPTY(inctx->bb))) {
            BIO_set_retry_read(bio);
            return -1;
        }

        if (block == APR_BLOCK_READ
            && APR_STATUS_IS_TIMEUP(inctx->rc)
            && APR_BRIGADE_EMPTY(inctx->bb)) {
            /* don't give up, just return the timeout */
            return -1;
        }
        if (inctx->rc != APR_SUCCESS) {
            /* Unexpected errors discard the brigade */
            apr_brigade_cleanup(inctx->bb);
            inctx->bb = NULL;
            return -1;
        }
    }

    inctx->rc = brigade_consume(inctx->bb, block, in, &inl);

    if (inctx->rc == APR_SUCCESS) {
        return (int)inl;
    }

    if (APR_STATUS_IS_EAGAIN(inctx->rc)
            || APR_STATUS_IS_EINTR(inctx->rc)) {
        BIO_set_retry_read(bio);
        return (int)inl;
    }

    /* Unexpected errors and APR_EOF clean out the brigade.
     * Subsequent calls will return APR_EOF.
     */
    apr_brigade_cleanup(inctx->bb);
    inctx->bb = NULL;

    if (APR_STATUS_IS_EOF(inctx->rc) && inl) {
        /* Provide the results of this read pass,
         * without resetting the BIO retry_read flag
         */
        return (int)inl;
    }

    return -1;
}

static int bio_filter_in_write(BIO *bio, const char *in, int inl)
{
    bio_filter_in_ctx_t *inctx = (bio_filter_in_ctx_t *)BIO_get_data(bio);
    ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, inctx->f->c,
                  "BUG: %s() should not be called", "bio_filter_in_write");
    AP_DEBUG_ASSERT(0);
    return -1;
}

static int bio_filter_in_puts(BIO *bio, const char *str)
{
    bio_filter_in_ctx_t *inctx = (bio_filter_in_ctx_t *)BIO_get_data(bio);
    ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, inctx->f->c,
                  "BUG: %s() should not be called", "bio_filter_in_puts");
    AP_DEBUG_ASSERT(0);
    return -1;
}

static int bio_filter_in_gets(BIO *bio, char *buf, int size)
{
    bio_filter_in_ctx_t *inctx = (bio_filter_in_ctx_t *)BIO_get_data(bio);
    ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, inctx->f->c,
                  "BUG: %s() should not be called", "bio_filter_in_gets");
    AP_DEBUG_ASSERT(0);
    return -1;
}

static long bio_filter_in_ctrl(BIO *bio, int cmd, long num, void *ptr)
{
    bio_filter_in_ctx_t *inctx = (bio_filter_in_ctx_t *)BIO_get_data(bio);
    switch (cmd) {
#ifdef BIO_CTRL_EOF
    case BIO_CTRL_EOF:
        return inctx->rc == APR_EOF;
#endif
    default:
        break;
    }
    ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, inctx->f->c,
                  "BUG: bio_filter_in_ctrl() should not be called with cmd=%i",
                  cmd);
    return 0;
}

#if MODSSL_USE_OPENSSL_PRE_1_1_API

static BIO_METHOD bio_filter_out_method = {
    BIO_TYPE_MEM,
    "APR output filter",
    bio_filter_out_write,
    bio_filter_out_read,     /* read is never called */
    bio_filter_out_puts,     /* puts is never called */
    bio_filter_out_gets,     /* gets is never called */
    bio_filter_out_ctrl,
    bio_filter_create,
    bio_filter_destroy,
    NULL
};

static BIO_METHOD bio_filter_in_method = {
    BIO_TYPE_MEM,
    "APR input filter",
    bio_filter_in_write,        /* write is never called */
    bio_filter_in_read,
    bio_filter_in_puts,         /* puts is never called */
    bio_filter_in_gets,         /* gets is never called */
    bio_filter_in_ctrl,         /* ctrl is called for EOF check */
    bio_filter_create,
    bio_filter_destroy,
    NULL
};

#else

static BIO_METHOD *bio_filter_out_method = NULL;
static BIO_METHOD *bio_filter_in_method = NULL;

void init_bio_methods(void)
{
    bio_filter_out_method = BIO_meth_new(BIO_TYPE_MEM, "APR output filter");
    BIO_meth_set_write(bio_filter_out_method, &bio_filter_out_write);
    BIO_meth_set_read(bio_filter_out_method, &bio_filter_out_read); /* read is never called */
    BIO_meth_set_puts(bio_filter_out_method, &bio_filter_out_puts); /* puts is never called */
    BIO_meth_set_gets(bio_filter_out_method, &bio_filter_out_gets); /* gets is never called */
    BIO_meth_set_ctrl(bio_filter_out_method, &bio_filter_out_ctrl);
    BIO_meth_set_create(bio_filter_out_method, &bio_filter_create);
    BIO_meth_set_destroy(bio_filter_out_method, &bio_filter_destroy);

    bio_filter_in_method = BIO_meth_new(BIO_TYPE_MEM, "APR input filter");
    BIO_meth_set_write(bio_filter_in_method, &bio_filter_in_write); /* write is never called */
    BIO_meth_set_read(bio_filter_in_method, &bio_filter_in_read);
    BIO_meth_set_puts(bio_filter_in_method, &bio_filter_in_puts);   /* puts is never called */
    BIO_meth_set_gets(bio_filter_in_method, &bio_filter_in_gets);   /* gets is never called */
    BIO_meth_set_ctrl(bio_filter_in_method, &bio_filter_in_ctrl);   /* ctrl is never called */
    BIO_meth_set_create(bio_filter_in_method, &bio_filter_create);
    BIO_meth_set_destroy(bio_filter_in_method, &bio_filter_destroy);
}

void free_bio_methods(void)
{
    BIO_meth_free(bio_filter_out_method);
    BIO_meth_free(bio_filter_in_method);
}
#endif

static apr_status_t ssl_io_input_read(bio_filter_in_ctx_t *inctx,
                                      char *buf,
                                      apr_size_t *len)
{
    apr_size_t wanted = *len;
    apr_size_t bytes = 0;
    int rc;

    *len = 0;

    /* If we have something leftover from last time, try that first. */
    if ((bytes = char_buffer_read(&inctx->cbuf, buf, wanted))) {
        *len = bytes;
        if (inctx->mode == AP_MODE_SPECULATIVE) {
            /* We want to rollback this read. */
            if (inctx->cbuf.length > 0) {
                inctx->cbuf.value -= bytes;
                inctx->cbuf.length += bytes;
            } else {
                char_buffer_write(&inctx->cbuf, buf, (int)bytes);
            }
            return APR_SUCCESS;
        }
        /* This could probably be *len == wanted, but be safe from stray
         * photons.
         */
        if (*len >= wanted) {
            return APR_SUCCESS;
        }
        if (inctx->mode == AP_MODE_GETLINE) {
            if (memchr(buf, APR_ASCII_LF, *len)) {
                return APR_SUCCESS;
            }
        }
        else {
            /* Down to a nonblock pattern as we have some data already
             */
            inctx->block = APR_NONBLOCK_READ;
        }
    }

    while (1) {

        if (!inctx->filter_ctx->pssl) {
            /* Ensure a non-zero error code is returned */
            if (inctx->rc == APR_SUCCESS) {
                inctx->rc = APR_EGENERAL;
            }
            break;
        }

        /* We rely on SSL_get_error() after the read, which requires an empty
         * error queue before the read in order to work properly.
         */
        ERR_clear_error();

        /* SSL_read may not read because we haven't taken enough data
         * from the stack.  This is where we want to consider all of
         * the blocking and SPECULATIVE semantics
         */
        rc = SSL_read(inctx->filter_ctx->pssl, buf + bytes, wanted - bytes);

        if (rc > 0) {
            *len += rc;
            if (inctx->mode == AP_MODE_SPECULATIVE) {
                /* We want to rollback this read. */
                char_buffer_write(&inctx->cbuf, buf, rc);
            }
            return inctx->rc;
        }
        else /* (rc <= 0) */ {
            int ssl_err;
            conn_rec *c;
            if (rc == 0) {
                /* If EAGAIN, we will loop given a blocking read,
                 * otherwise consider ourselves at EOF.
                 */
                if (APR_STATUS_IS_EAGAIN(inctx->rc)
                        || APR_STATUS_IS_EINTR(inctx->rc)) {
                    /* Already read something, return APR_SUCCESS instead.
                     * On win32 in particular, but perhaps on other kernels,
                     * a blocking call isn't 'always' blocking.
                     */
                    if (*len > 0) {
                        inctx->rc = APR_SUCCESS;
                        break;
                    }
                    if (inctx->block == APR_NONBLOCK_READ) {
                        break;
                    }
                }
                else {
                    if (*len > 0) {
                        inctx->rc = APR_SUCCESS;
                        break;
                    }
                }
            }
            ssl_err = SSL_get_error(inctx->filter_ctx->pssl, rc);
            c = (conn_rec*)SSL_get_app_data(inctx->filter_ctx->pssl);

            if (ssl_err == SSL_ERROR_WANT_READ) {
                /*
                 * If OpenSSL wants to read more, and we were nonblocking,
                 * report as an EAGAIN.  Otherwise loop, pulling more
                 * data from network filter.
                 *
                 * (This is usually the case when the client forces an SSL
                 * renegotiation which is handled implicitly by OpenSSL.)
                 */
                inctx->rc = APR_EAGAIN;

                if (*len > 0) {
                    inctx->rc = APR_SUCCESS;
                    break;
                }
                if (inctx->block == APR_NONBLOCK_READ) {
                    break;
                }
                continue;  /* Blocking and nothing yet?  Try again. */
            }
            else if (ssl_err == SSL_ERROR_SYSCALL) {
                if (APR_STATUS_IS_EAGAIN(inctx->rc)
                        || APR_STATUS_IS_EINTR(inctx->rc)) {
                    /* Already read something, return APR_SUCCESS instead. */
                    if (*len > 0) {
                        inctx->rc = APR_SUCCESS;
                        break;
                    }
                    if (inctx->block == APR_NONBLOCK_READ) {
                        break;
                    }
                    continue;  /* Blocking and nothing yet?  Try again. */
                }
                else if (APR_STATUS_IS_TIMEUP(inctx->rc)) {
                    /* just return it, the calling layer might be fine with it,
                       and we do not want to bloat the log. */
                }
                else {
                    ap_log_cerror(APLOG_MARK, APLOG_INFO, inctx->rc, c, APLOGNO(01991)
                                  "SSL input filter read failed.");
                }
            }
            else if (rc == 0 && ssl_err == SSL_ERROR_ZERO_RETURN) {
                inctx->rc = APR_EOF;
                break;
            }
            else /* if (ssl_err == SSL_ERROR_SSL) */ {
                /*
                 * Log SSL errors and any unexpected conditions.
                 */
                ap_log_cerror(APLOG_MARK, APLOG_INFO, inctx->rc, c, APLOGNO(01992)
                              "SSL library error %d reading data", ssl_err);
                ssl_log_ssl_error(SSLLOG_MARK, APLOG_INFO, mySrvFromConn(c));

            }
            if (rc == 0) {
                inctx->rc = APR_EOF;
                break;
            }
            if (inctx->rc == APR_SUCCESS) {
                inctx->rc = APR_EGENERAL;
            }
            break;
        }
    }
    return inctx->rc;
}

/* Read a line of input from the SSL input layer into buffer BUF of
* length *LEN; updating *len to reflect the length of the line
* including the LF character. */
static apr_status_t ssl_io_input_getline(bio_filter_in_ctx_t *inctx,
                                         char *buf,
                                         apr_size_t *len)
{
    const char *pos = NULL;
    apr_status_t status;
    apr_size_t tmplen = *len, buflen = *len, offset = 0;

    *len = 0;

    /*
     * in most cases we get all the headers on the first SSL_read.
     * however, in certain cases SSL_read will only get a partial
     * chunk of the headers, so we try to read until LF is seen.
     */

    while (tmplen > 0) {
        status = ssl_io_input_read(inctx, buf + offset, &tmplen);

        if (status != APR_SUCCESS) {
            if (APR_STATUS_IS_EAGAIN(status) && (*len > 0)) {
                /* Save the part of the line we already got */
                char_buffer_write(&inctx->cbuf, buf, *len);
            }
            return status;
        }

        *len += tmplen;

        if ((pos = memchr(buf, APR_ASCII_LF, *len))) {
            break;
        }

        offset += tmplen;
        tmplen = buflen - offset;
    }

    if (pos) {
        char *value;
        int length;
        apr_size_t bytes = pos - buf;

        bytes += 1;
        value = buf + bytes;
        length = *len - bytes;

        char_buffer_write(&inctx->cbuf, value, length);

        *len = bytes;
    }

    return APR_SUCCESS;
}

static apr_status_t ssl_filter_write(ap_filter_t *f,
                                     const char *data,
                                     apr_size_t len)
{
    ssl_filter_ctx_t *filter_ctx = f->ctx;
    bio_filter_out_ctx_t *outctx;
    int res;

    /* write SSL */
    if (filter_ctx->pssl == NULL) {
        return APR_EGENERAL;
    }

    ap_log_cerror(APLOG_MARK, APLOG_TRACE6, 0, f->c,
                  "ssl_filter_write: %"APR_SIZE_T_FMT" bytes", len);

    /* We rely on SSL_get_error() after the write, which requires an empty error
     * queue before the write in order to work properly.
     */
    ERR_clear_error();

    outctx = (bio_filter_out_ctx_t *)BIO_get_data(filter_ctx->pbioWrite);
    res = SSL_write(filter_ctx->pssl, (unsigned char *)data, len);

    if (res < 0) {
        int ssl_err = SSL_get_error(filter_ctx->pssl, res);
        conn_rec *c = (conn_rec*)SSL_get_app_data(outctx->filter_ctx->pssl);

        if (ssl_err == SSL_ERROR_WANT_WRITE) {
            /*
             * If OpenSSL wants to write more, and we were nonblocking,
             * report as an EAGAIN.  Otherwise loop, pushing more
             * data at the network filter.
             *
             * (This is usually the case when the client forces an SSL
             * renegotiation which is handled implicitly by OpenSSL.)
             */
            outctx->rc = APR_EAGAIN;
        }
        else if (ssl_err == SSL_ERROR_WANT_READ) {
            /*
             * If OpenSSL wants to read during write, and we were
             * nonblocking, set the sense explicitly to read and
             * report as an EAGAIN.
             *
             * (This is usually the case when the client forces an SSL
             * renegotiation which is handled implicitly by OpenSSL.)
             */
            outctx->c->cs->sense = CONN_SENSE_WANT_READ;
            outctx->rc = APR_EAGAIN;
            ap_log_cerror(APLOG_MARK, APLOG_TRACE6, 0, outctx->c,
                          "Want read during nonblocking write");
        }
        else if (ssl_err == SSL_ERROR_SYSCALL) {
            ap_log_cerror(APLOG_MARK, APLOG_INFO, outctx->rc, c, APLOGNO(01993)
                          "SSL output filter write failed.");
        }
        else /* if (ssl_err == SSL_ERROR_SSL) */ {
            /*
             * Log SSL errors
             */
            ap_log_cerror(APLOG_MARK, APLOG_INFO, outctx->rc, c, APLOGNO(01994)
                          "SSL library error %d writing data", ssl_err);
            ssl_log_ssl_error(SSLLOG_MARK, APLOG_INFO, mySrvFromConn(c));
        }
        if (outctx->rc == APR_SUCCESS) {
            outctx->rc = APR_EGENERAL;
        }
    }
    else if ((apr_size_t)res != len) {
        conn_rec *c = f->c;
        char *reason = "reason unknown";

        /* XXX: probably a better way to determine this */
        if (SSL_total_renegotiations(filter_ctx->pssl)) {
            reason = "likely due to failed renegotiation";
        }

        ap_log_cerror(APLOG_MARK, APLOG_INFO, outctx->rc, c, APLOGNO(01995)
                      "failed to write %" APR_SSIZE_T_FMT
                      " of %" APR_SIZE_T_FMT " bytes (%s)",
                      len - (apr_size_t)res, len, reason);

        outctx->rc = APR_EGENERAL;
    }
    return outctx->rc;
}

/* Just use a simple request.  Any request will work for this, because
* we use a flag in the conn_rec->conn_vector now.  The fake request just
* gets the request back to the Apache core so that a response can be sent.
* Since we use an HTTP/1.x request, we also have to inject the empty line
* that terminates the headers, or the core will read more data from the
* socket.
*/
#define HTTP_ON_HTTPS_PORT \
    "GET / HTTP/1.0" CRLF

#define HTTP_ON_HTTPS_PORT_BUCKET(alloc) \
    apr_bucket_immortal_create(HTTP_ON_HTTPS_PORT, \
                               sizeof(HTTP_ON_HTTPS_PORT) - 1, \
                               alloc)

/* Custom apr_status_t error code, used when a plain HTTP request is
* received on an SSL port. */
#define MODSSL_ERROR_HTTP_ON_HTTPS (APR_OS_START_USERERR + 0)

/* Custom apr_status_t error code, used when the proxy cannot
* establish an outgoing SSL connection. */
#define MODSSL_ERROR_BAD_GATEWAY (APR_OS_START_USERERR + 1)

static void ssl_io_filter_disable(SSLConnRec *sslconn,
                                  bio_filter_in_ctx_t *inctx)
{
    SSL_free(inctx->ssl);
    sslconn->ssl = NULL;
    inctx->ssl = NULL;
    inctx->filter_ctx->pssl = NULL;
}

static apr_status_t ssl_io_filter_error(bio_filter_in_ctx_t *inctx,
                                        apr_bucket_brigade *bb,
                                        apr_status_t status,
                                        int is_init)
{
    ap_filter_t *f = inctx->f;
    SSLConnRec *sslconn = myConnConfig(f->c);
    apr_bucket *bucket;
    int send_eos = 1;

    switch (status) {
    case MODSSL_ERROR_HTTP_ON_HTTPS:
            /* log the situation */
            ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, f->c, APLOGNO(01996)
                         "SSL handshake failed: HTTP spoken on HTTPS port; "
                         "trying to send HTML error page");
            ssl_log_ssl_error(SSLLOG_MARK, APLOG_INFO, sslconn->server);

            ssl_io_filter_disable(sslconn, inctx);
            f->c->keepalive = AP_CONN_CLOSE;
            if (is_init) {
                sslconn->non_ssl_request = NON_SSL_SEND_REQLINE;
                return APR_EGENERAL;
            }
            sslconn->non_ssl_request = NON_SSL_SEND_HDR_SEP;

            /* fake the request line */
            bucket = HTTP_ON_HTTPS_PORT_BUCKET(f->c->bucket_alloc);
            send_eos = 0;
            break;

    case MODSSL_ERROR_BAD_GATEWAY:
        ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, f->c, APLOGNO(01997)
                      "SSL handshake failed: sending 502");
        f->c->aborted = 1;
        return APR_EGENERAL;

    default:
        return status;
    }

    APR_BRIGADE_INSERT_TAIL(bb, bucket);
    if (send_eos) {
        bucket = apr_bucket_eos_create(f->c->bucket_alloc);
        APR_BRIGADE_INSERT_TAIL(bb, bucket);
    }
    return APR_SUCCESS;
}

static const char ssl_io_filter[] = "SSL/TLS Filter";
static const char ssl_io_buffer[] = "SSL/TLS Buffer";
static const char ssl_io_coalesce[] = "SSL/TLS Coalescing Filter";

/*
*  Close the SSL part of the socket connection
*  (called immediately _before_ the socket is closed)
*  or called with
*/
static void ssl_filter_io_shutdown(ssl_filter_ctx_t *filter_ctx,
                                   conn_rec *c, int abortive)
{
    SSL *ssl = filter_ctx->pssl;
    const char *type = "";
    SSLConnRec *sslconn = myConnConfig(c);
    int shutdown_type;
    int loglevel = APLOG_DEBUG;
    const char *logno;

    if (!ssl) {
        return;
    }

    /*
     * Now close the SSL layer of the connection. We've to take
     * the TLSv1 standard into account here:
     *
     * | 7.2.1. Closure alerts
     * |
     * | The client and the server must share knowledge that the connection is
     * | ending in order to avoid a truncation attack. Either party may
     * | initiate the exchange of closing messages.
     * |
     * | close_notify
     * |     This message notifies the recipient that the sender will not send
     * |     any more messages on this connection. The session becomes
     * |     unresumable if any connection is terminated without proper
     * |     close_notify messages with level equal to warning.
     * |
     * | Either party may initiate a close by sending a close_notify alert.
     * | Any data received after a closure alert is ignored.
     * |
     * | Each party is required to send a close_notify alert before closing
     * | the write side of the connection. It is required that the other party
     * | respond with a close_notify alert of its own and close down the
     * | connection immediately, discarding any pending writes. It is not
     * | required for the initiator of the close to wait for the responding
     * | close_notify alert before closing the read side of the connection.
     *
     * This means we've to send a close notify message, but haven't to wait
     * for the close notify of the client. Actually we cannot wait for the
     * close notify of the client because some clients (including Netscape
     * 4.x) don't send one, so we would hang.
     */

    /*
     * exchange close notify messages, but allow the user
     * to force the type of handshake via SetEnvIf directive
     */
    if (abortive) {
        shutdown_type = SSL_SENT_SHUTDOWN|SSL_RECEIVED_SHUTDOWN;
        type = "abortive";
        logno = APLOGNO(01998);
        loglevel = APLOG_INFO;
    }
    else switch (sslconn->shutdown_type) {
      case SSL_SHUTDOWN_TYPE_UNCLEAN:
        /* perform no close notify handshake at all
           (violates the SSL/TLS standard!) */
        shutdown_type = SSL_SENT_SHUTDOWN|SSL_RECEIVED_SHUTDOWN;
        type = "unclean";
        logno = APLOGNO(01999);
        break;
      case SSL_SHUTDOWN_TYPE_ACCURATE:
        /* send close notify and wait for clients close notify
           (standard compliant, but usually causes connection hangs) */
        shutdown_type = 0;
        type = "accurate";
        logno = APLOGNO(02000);
        break;
      default:
        /*
         * case SSL_SHUTDOWN_TYPE_UNSET:
         * case SSL_SHUTDOWN_TYPE_STANDARD:
         */
        /* send close notify, but don't wait for clients close notify
           (standard compliant and safe, so it's the DEFAULT!) */
        shutdown_type = SSL_RECEIVED_SHUTDOWN;
        type = "standard";
        logno = APLOGNO(02001);
        break;
    }

    SSL_set_shutdown(ssl, shutdown_type);
    modssl_smart_shutdown(ssl);

    /* and finally log the fact that we've closed the connection */
    if (APLOG_CS_IS_LEVEL(c, mySrvFromConn(c), loglevel)) {
        /* Intentional no APLOGNO */
        /* logno provides APLOGNO */
        ap_log_cserror(APLOG_MARK, loglevel, 0, c, mySrvFromConn(c),
                       "%sConnection closed to child %ld with %s shutdown "
                       "(server %s)",
                       logno, c->id, type,
                       ssl_util_vhostid(c->pool, mySrvFromConn(c)));
    }

    /* deallocate the SSL connection */
    if (sslconn->client_cert) {
        X509_free(sslconn->client_cert);
        sslconn->client_cert = NULL;
    }
    SSL_free(ssl);
    sslconn->ssl = NULL;
    filter_ctx->pssl = NULL; /* so filters know we've been shutdown */

    if (abortive) {
        /* prevent any further I/O */
        c->aborted = 1;
    }
}

static apr_status_t ssl_io_filter_cleanup(void *data)
{
    ssl_filter_ctx_t *filter_ctx = data;

    if (filter_ctx->pssl) {
        conn_rec *c = (conn_rec *)SSL_get_app_data(filter_ctx->pssl);
        SSLConnRec *sslconn = myConnConfig(c);

        SSL_free(filter_ctx->pssl);
        sslconn->ssl = filter_ctx->pssl = NULL;
    }

    return APR_SUCCESS;
}

/*
* The hook is NOT registered with ap_hook_process_connection. Instead, it is
* called manually from the churn () before it tries to read any data.
* There is some problem if I accept conn_rec *. Still investigating..
* Adv. if conn_rec * can be accepted is we can hook this function using the
* ap_hook_process_connection hook.
*/

/* Perform the SSL handshake (whether in client or server mode), if
* necessary, for the given connection. */
static apr_status_t ssl_io_filter_handshake(ssl_filter_ctx_t *filter_ctx)
{
    conn_rec *c         = (conn_rec *)SSL_get_app_data(filter_ctx->pssl);
    SSLConnRec *sslconn = myConnConfig(c);
    SSLSrvConfigRec *sc;
    X509 *cert;
    int n;
    int ssl_err;
    long verify_result;
    server_rec *server;

    if (SSL_is_init_finished(filter_ctx->pssl)) {
        return APR_SUCCESS;
    }

    server = sslconn->server;
    if (c->outgoing) {
#ifdef HAVE_TLSEXT
        apr_ipsubnet_t *ip;
#ifdef HAVE_TLS_ALPN
        const char *alpn_note;
        apr_array_header_t *alpn_proposed = NULL;
        int alpn_empty_ok = 1;
#endif
#endif
        const char *hostname_note = apr_table_get(c->notes,
                                                  "proxy-request-hostname");
        BOOL proxy_ssl_check_peer_ok = TRUE;
        int post_handshake_rc = OK;
        SSLDirConfigRec *dc;

        dc = sslconn->dc;
        sc = mySrvConfig(server);

#ifdef HAVE_TLSEXT
#ifdef HAVE_TLS_ALPN
        alpn_note = apr_table_get(c->notes, "proxy-request-alpn-protos");
        if (alpn_note) {
            char *protos, *s, *p, *last, *proto;
            apr_size_t len;

            /* Transform the note into a protocol formatted byte array:
             * (len-byte proto-char+)*
             * We need the remote server to agree on one of these, unless 'http/1.1'
             * is also among our proposals. Because pre-ALPN remotes will speak this.
             */
            alpn_proposed = apr_array_make(c->pool, 3, sizeof(const char*));
            alpn_empty_ok = 0;
            s = protos = apr_pcalloc(c->pool, strlen(alpn_note)+1);
            p = apr_pstrdup(c->pool, alpn_note);
            while ((p = apr_strtok(p, ", ", &last))) {
                len = last - p - (*last? 1 : 0);
                if (len > 255) {
                    ap_log_cerror(APLOG_MARK, APLOG_ERR, 0, c, APLOGNO(03309)
                                  "ALPN proxy protocol identifier too long: %s",
                                  p);
                    ssl_log_ssl_error(SSLLOG_MARK, APLOG_ERR, server);
                    return APR_EGENERAL;
                }
                proto = apr_pstrndup(c->pool, p, len);
                APR_ARRAY_PUSH(alpn_proposed, const char*) = proto;
                if (!strcmp("http/1.1", proto)) {
                    alpn_empty_ok = 1;
                }
                *s++ = (unsigned char)len;
                while (len--) {
                    *s++ = *p++;
                }
                p = NULL;
            }
            ap_log_cerror(APLOG_MARK, APLOG_TRACE1, 0, c,
                          "setting alpn protos from '%s', protolen=%d",
                          alpn_note, (int)(s - protos));
            if (protos != s && SSL_set_alpn_protos(filter_ctx->pssl,
                                                   (unsigned char *)protos,
                                                   s - protos)) {
                ap_log_cerror(APLOG_MARK, APLOG_WARNING, 0, c, APLOGNO(03310)
                              "error setting alpn protos from '%s'", alpn_note);
                ssl_log_ssl_error(SSLLOG_MARK, APLOG_WARNING, server);
                /* If ALPN was requested and we cannot do it, we must fail */
                return MODSSL_ERROR_BAD_GATEWAY;
            }
        }
#endif /* defined HAVE_TLS_ALPN */
        /*
         * Enable SNI for backend requests. Make sure we don't do it for
         * pure SSLv3 connections, and also prevent IP addresses
         * from being included in the SNI extension. (OpenSSL would simply
         * pass them on, but RFC 6066 is quite clear on this: "Literal
         * IPv4 and IPv6 addresses are not permitted".)
         */
        if (hostname_note &&
#ifndef OPENSSL_NO_SSL3
            dc->proxy->protocol != SSL_PROTOCOL_SSLV3 &&
#endif
            apr_ipsubnet_create(&ip, hostname_note, NULL,
                                c->pool) != APR_SUCCESS) {
            if (SSL_set_tlsext_host_name(filter_ctx->pssl, hostname_note)) {
                ap_log_cerror(APLOG_MARK, APLOG_TRACE3, 0, c,
                              "SNI extension for SSL Proxy request set to '%s'",
                              hostname_note);
            } else {
                ap_log_cerror(APLOG_MARK, APLOG_WARNING, 0, c, APLOGNO(02002)
                              "Failed to set SNI extension for SSL Proxy "
                              "request to '%s'", hostname_note);
                ssl_log_ssl_error(SSLLOG_MARK, APLOG_WARNING, server);
            }
        }
#endif /* defined HAVE_TLSEXT */

        if ((n = SSL_connect(filter_ctx->pssl)) <= 0) {
            ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(02003)
                          "SSL Proxy connect failed");
            ssl_log_ssl_error(SSLLOG_MARK, APLOG_INFO, server);
            /* ensure that the SSL structures etc are freed, etc: */
            ssl_filter_io_shutdown(filter_ctx, c, 1);
            apr_table_setn(c->notes, "SSL_connect_rv", "err");
            return MODSSL_ERROR_BAD_GATEWAY;
        }

        cert = SSL_get_peer_certificate(filter_ctx->pssl);

        if (dc->proxy->ssl_check_peer_expire != FALSE) {
            if (!cert
                || (X509_cmp_current_time(
                     X509_get_notBefore(cert)) >= 0)
                || (X509_cmp_current_time(
                     X509_get_notAfter(cert)) <= 0)) {
                proxy_ssl_check_peer_ok = FALSE;
                ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(02004)
                              "SSL Proxy: Peer certificate is expired");
            }
        }
        if ((dc->proxy->ssl_check_peer_name != FALSE) &&
            ((dc->proxy->ssl_check_peer_cn != FALSE) ||
             (dc->proxy->ssl_check_peer_name == TRUE)) &&
            hostname_note) {
            if (!cert
                || modssl_X509_match_name(c->pool, cert, hostname_note,
                                          TRUE, server) == FALSE) {
                proxy_ssl_check_peer_ok = FALSE;
                ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(02411)
                              "SSL Proxy: Peer certificate does not match "
                              "for hostname %s", hostname_note);
            }
        }
        else if ((dc->proxy->ssl_check_peer_cn == TRUE) &&
            hostname_note) {
            const char *hostname;
            int match = 0;

            hostname = ssl_var_lookup(NULL, server, c, NULL,
                                      "SSL_CLIENT_S_DN_CN");

            /* Do string match or simplest wildcard match if that
             * fails. */
            match = strcasecmp(hostname, hostname_note) == 0;
            if (!match && strncmp(hostname, "*.", 2) == 0) {
                const char *p = ap_strchr_c(hostname_note, '.');

                match = p && strcasecmp(p, hostname + 1) == 0;
            }

            if (!match) {
                proxy_ssl_check_peer_ok = FALSE;
                ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(02005)
                              "SSL Proxy: Peer certificate CN mismatch:"
                              " Certificate CN: %s Requested hostname: %s",
                              hostname, hostname_note);
            }
        }

#ifdef HAVE_TLS_ALPN
        /* If we proposed ALPN protocol(s), we need to check if the server
         * agreed to one of them. While <https://www.rfc-editor.org/rfc/rfc7301.txt>
         * chapter 3.2 says the server SHALL error the handshake in such a case,
         * the reality is that some servers fall back to their default, e.g. http/1.1.
         * (we also do this right now)
         * We need to treat this as an error for security reasons.
         */
        if (alpn_proposed && alpn_proposed->nelts > 0) {
            const char *selected;
            unsigned int slen;

            SSL_get0_alpn_selected(filter_ctx->pssl, (const unsigned char**)&selected, &slen);
            if (!selected || !slen) {
                /* No ALPN selection reported by the remote server. This could mean
                 * it does not support ALPN (old server) or that it does not support
                 * any of our proposals (Apache itself up to 2.4.48 at least did that). */
               if (!alpn_empty_ok) {
                    ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(10273)
                                  "SSL Proxy: Peer did not select any of our ALPN protocols [%s].",
                                  alpn_note);
                    proxy_ssl_check_peer_ok = FALSE;
               }
            }
            else {
                const char *proto;
                int i, found = 0;
                for (i = 0; !found && i < alpn_proposed->nelts; ++i) {
                    proto = APR_ARRAY_IDX(alpn_proposed, i, const char *);
                    found = !strncmp(selected, proto, slen);
                }
                if (!found) {
                    /* From a conforming peer, this should never happen,
                     * but life always finds a way... */
                    proto = apr_pstrndup(c->pool, selected, slen);
                    ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(10274)
                                  "SSL Proxy: Peer proposed ALPN protocol %s which is none "
                                  "of our proposals [%s].", proto, alpn_note);
                    proxy_ssl_check_peer_ok = FALSE;
                }
            }
        }
#endif

        if (proxy_ssl_check_peer_ok == TRUE) {
            /* another chance to fail */
            post_handshake_rc = ssl_run_proxy_post_handshake(c, filter_ctx->pssl);
        }

        if (cert) {
            X509_free(cert);
        }

        if (proxy_ssl_check_peer_ok != TRUE
            || (post_handshake_rc != OK && post_handshake_rc != DECLINED)) {
            /* ensure that the SSL structures etc are freed, etc: */
            ssl_filter_io_shutdown(filter_ctx, c, 1);
            apr_table_setn(c->notes, "SSL_connect_rv", "err");
            return MODSSL_ERROR_BAD_GATEWAY;
        }

        apr_table_setn(c->notes, "SSL_connect_rv", "ok");
        return APR_SUCCESS;
    }

    /* We rely on SSL_get_error() after the accept, which requires an empty
     * error queue before the accept in order to work properly.
     */
    ERR_clear_error();

    if ((n = SSL_accept(filter_ctx->pssl)) <= 0) {
        bio_filter_in_ctx_t *inctx = (bio_filter_in_ctx_t *)
                                     BIO_get_data(filter_ctx->pbioRead);
        bio_filter_out_ctx_t *outctx = (bio_filter_out_ctx_t *)
                                       BIO_get_data(filter_ctx->pbioWrite);
        apr_status_t rc = inctx->rc ? inctx->rc : outctx->rc ;
        ssl_err = SSL_get_error(filter_ctx->pssl, n);

        if (ssl_err == SSL_ERROR_ZERO_RETURN) {
            /*
             * The case where the connection was closed before any data
             * was transferred. That's not a real error and can occur
             * sporadically with some clients.
             */
            ap_log_cerror(APLOG_MARK, APLOG_INFO, rc, c, APLOGNO(02006)
                         "SSL handshake stopped: connection was closed");
        }
        else if (ssl_err == SSL_ERROR_WANT_READ) {
            /*
             * This is in addition to what was present earlier. It is
             * borrowed from openssl_state_machine.c [mod_tls].
             * TBD.
             */
            outctx->rc = APR_EAGAIN;
            return APR_EAGAIN;
        }
        else if (ERR_GET_LIB(ERR_peek_error()) == ERR_LIB_SSL &&
                 ERR_GET_REASON(ERR_peek_error()) == SSL_R_HTTP_REQUEST) {
            /*
             * The case where OpenSSL has recognized a HTTP request:
             * This means the client speaks plain HTTP on our HTTPS port.
             * ssl_io_filter_error will disable the ssl filters when it
             * sees this status code.
             */
            return MODSSL_ERROR_HTTP_ON_HTTPS;
        }
        else if (ssl_err == SSL_ERROR_SYSCALL) {
            ap_log_cerror(APLOG_MARK, APLOG_DEBUG, rc, c, APLOGNO(02007)
                          "SSL handshake interrupted by system "
                          "[Hint: Stop button pressed in browser?!]");
        }
        else /* if (ssl_err == SSL_ERROR_SSL) */ {
            /*
             * Log SSL errors and any unexpected conditions.
             */
            ap_log_cerror(APLOG_MARK, APLOG_INFO, rc, c, APLOGNO(02008)
                          "SSL library error %d in handshake "
                          "(server %s)", ssl_err,
                          ssl_util_vhostid(c->pool, server));
            ssl_log_ssl_error(SSLLOG_MARK, APLOG_INFO, server);

        }
        if (inctx->rc == APR_SUCCESS) {
            inctx->rc = APR_EGENERAL;
        }

        ssl_filter_io_shutdown(filter_ctx, c, 1);
        return inctx->rc;
    }
    sc = mySrvConfig(sslconn->server);

    /*
     * Check for failed client authentication
     */
    verify_result = SSL_get_verify_result(filter_ctx->pssl);

    if ((verify_result != X509_V_OK) ||
        sslconn->verify_error)
    {
        if (ssl_verify_error_is_optional(verify_result) &&
            (sc->server->auth.verify_mode == SSL_CVERIFY_OPTIONAL_NO_CA))
        {
            /* leaving this log message as an error for the moment,
             * according to the mod_ssl docs:
             * "level optional_no_ca is actually against the idea
             *  of authentication (but can be used to establish
             * SSL test pages, etc.)"
             * optional_no_ca doesn't appear to work as advertised
             * in 1.x
             */
            ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(02009)
                          "SSL client authentication failed, "
                          "accepting certificate based on "
                          "\"SSLVerifyClient optional_no_ca\" "
                          "configuration");
            ssl_log_ssl_error(SSLLOG_MARK, APLOG_INFO, server);
        }
        else {
            const char *error = sslconn->verify_error ?
                sslconn->verify_error :
                X509_verify_cert_error_string(verify_result);

            ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(02010)
                         "SSL client authentication failed: %s",
                         error ? error : "unknown");
            ssl_log_ssl_error(SSLLOG_MARK, APLOG_INFO, server);

            ssl_filter_io_shutdown(filter_ctx, c, 1);
            return APR_ECONNABORTED;
        }
    }

    /*
     * Remember the peer certificate's DN
     */
    if ((cert = SSL_get_peer_certificate(filter_ctx->pssl))) {
        if (sslconn->client_cert) {
            X509_free(sslconn->client_cert);
        }
        sslconn->client_cert = cert;
        sslconn->client_dn = NULL;
    }

    /*
     * Make really sure that when a peer certificate
     * is required we really got one... (be paranoid)
     */
    if ((sc->server->auth.verify_mode == SSL_CVERIFY_REQUIRE) &&
        !sslconn->client_cert)
    {
        ap_log_cerror(APLOG_MARK, APLOG_INFO, 0, c, APLOGNO(02011)
                      "No acceptable peer certificate available");

        ssl_filter_io_shutdown(filter_ctx, c, 1);
        return APR_ECONNABORTED;
    }

    return APR_SUCCESS;
}

static apr_status_t ssl_io_filter_input(ap_filter_t *f,
                                        apr_bucket_brigade *bb,
                                        ap_input_mode_t mode,
                                        apr_read_type_e block,
                                        apr_off_t readbytes)
{
    apr_status_t status;
    bio_filter_in_ctx_t *inctx = f->ctx;
    const char *start = inctx->buffer; /* start of block to return */
    apr_size_t len = sizeof(inctx->buffer); /* length of block to return */
    int is_init = (mode == AP_MODE_INIT);
    apr_bucket *bucket;

    if (f->c->aborted) {
        /* XXX: Ok, if we aborted, we ARE at the EOS.  We also have
         * aborted.  This 'double protection' is probably redundant,
         * but also effective against just about anything.
         */
        bucket = apr_bucket_eos_create(f->c->bucket_alloc);
        APR_BRIGADE_INSERT_TAIL(bb, bucket);
        return APR_ECONNABORTED;
    }

    if (!inctx->ssl) {
        SSLConnRec *sslconn = myConnConfig(f->c);
        if (sslconn->non_ssl_request == NON_SSL_SEND_REQLINE) {
            bucket = HTTP_ON_HTTPS_PORT_BUCKET(f->c->bucket_alloc);
            APR_BRIGADE_INSERT_TAIL(bb, bucket);
            if (mode != AP_MODE_SPECULATIVE) {
                sslconn->non_ssl_request = NON_SSL_SEND_HDR_SEP;
            }
            return APR_SUCCESS;
        }
        if (sslconn->non_ssl_request == NON_SSL_SEND_HDR_SEP) {
            bucket = apr_bucket_immortal_create(CRLF, 2, f->c->bucket_alloc);
            APR_BRIGADE_INSERT_TAIL(bb, bucket);
            if (mode != AP_MODE_SPECULATIVE) {
                sslconn->non_ssl_request = NON_SSL_SET_ERROR_MSG;
            }
            return APR_SUCCESS;
        }
        return ap_get_brigade(f->next, bb, mode, block, readbytes);
    }

    /* XXX: we don't currently support anything other than these modes. */
    if (mode != AP_MODE_READBYTES && mode != AP_MODE_GETLINE &&
        mode != AP_MODE_SPECULATIVE && mode != AP_MODE_INIT) {
        return APR_ENOTIMPL;
    }

    inctx->mode = mode;
    inctx->block = block;

    /* XXX: we could actually move ssl_io_filter_handshake to an
     * ap_hook_process_connection but would still need to call it for
     * AP_MODE_INIT for protocols that may upgrade the connection
     * rather than have SSLEngine On configured.
     */
    if ((status = ssl_io_filter_handshake(inctx->filter_ctx)) != APR_SUCCESS) {
        return ssl_io_filter_error(inctx, bb, status, is_init);
    }

    if (is_init) {
        /* protocol module needs to handshake before sending
         * data to client (e.g. NNTP or FTP)
         */
        return APR_SUCCESS;
    }

    if (inctx->mode == AP_MODE_READBYTES ||
        inctx->mode == AP_MODE_SPECULATIVE) {
        /* Protected from truncation, readbytes < MAX_SIZE_T
         * FIXME: No, it's *not* protected.  -- jre */
        if (readbytes < len) {
            len = (apr_size_t)readbytes;
        }
        status = ssl_io_input_read(inctx, inctx->buffer, &len);
    }
    else if (inctx->mode == AP_MODE_GETLINE) {
        const char *pos;

        /* Satisfy the read directly out of the buffer if possible;
         * invoking ssl_io_input_getline will mean the entire buffer
         * is copied once (unnecessarily) for each GETLINE call. */
        if (inctx->cbuf.length
            && (pos = memchr(inctx->cbuf.value, APR_ASCII_LF,
                             inctx->cbuf.length)) != NULL) {
            start = inctx->cbuf.value;
            len = 1 + pos - start; /* +1 to include LF */
            /* Buffer contents now consumed. */
            inctx->cbuf.value += len;
            inctx->cbuf.length -= len;
            status = APR_SUCCESS;
        }
        else {
            /* Otherwise fall back to the hard way. */
            status = ssl_io_input_getline(inctx, inctx->buffer, &len);
        }
    }
    else {
        /* We have no idea what you are talking about, so return an error. */
        status = APR_ENOTIMPL;
    }

    /* It is possible for mod_ssl's BIO to be used outside of the
     * direct control of mod_ssl's input or output filter -- notably,
     * when mod_ssl initiates a renegotiation.  Switching the BIO mode
     * back to "blocking" here ensures such operations don't fail with
     * SSL_ERROR_WANT_READ. */
    inctx->block = APR_BLOCK_READ;

    /* Handle custom errors. */
    if (status != APR_SUCCESS) {
        return ssl_io_filter_error(inctx, bb, status, 0);
    }

    /* Create a transient bucket out of the decrypted data. */
    if (len > 0) {
        bucket =
            apr_bucket_transient_create(start, len, f->c->bucket_alloc);
        APR_BRIGADE_INSERT_TAIL(bb, bucket);
    }

    return APR_SUCCESS;
}

/* ssl_io_filter_output() produces one SSL/TLS record per bucket
* passed down the output filter stack.  This results in a high
* overhead (more network packets & TLS processing) for any output
* comprising many small buckets.  SSI output passed through the HTTP
* chunk filter, for example, may produce many brigades containing
* small buckets - [chunk-size CRLF] [chunk-data] [CRLF].
*
* Sending HTTP response headers as a separate TLS record to the
* response body also reveals information to a network observer (the
* size of headers) which can be significant.
*
* The coalescing filter merges data buckets with the aim of producing
* fewer, larger TLS records - without copying/buffering all content
* and introducing unnecessary overhead.
*
* ### This buffering could be probably be done more comprehensively
* ### in ssl_io_filter_output itself.
*
* ### Another possible performance optimisation in particular for the
* ### [HEAP] [FILE] HTTP response case is using a brigade rather than
* ### a char array to buffer; using apr_brigade_write() to append
* ### will use already-allocated memory from the HEAP, reducing # of
* ### copies.
*/

#define COALESCE_BYTES (AP_IOBUFSIZE)

struct coalesce_ctx {
    char buffer[COALESCE_BYTES];
    apr_size_t bytes; /* number of bytes of buffer used. */
};

static apr_status_t ssl_io_filter_coalesce(ap_filter_t *f,
                                           apr_bucket_brigade *bb)
{
    apr_bucket *e, *upto;
    apr_size_t bytes = 0;
    struct coalesce_ctx *ctx = f->ctx;
    apr_size_t buffered = ctx ? ctx->bytes : 0; /* space used on entry */
    unsigned count = 0;

    /* The brigade consists of zero-or-more small data buckets which
     * can be coalesced (referred to as the "prefix"), followed by the
     * remainder of the brigade.
     *
     * Find the last bucket - if any - of that prefix.  count gives
     * the number of buckets in the prefix.  The "prefix" must contain
     * only data buckets with known length, and must be of a total
     * size which fits into the buffer.
     *
     * N.B.: The process here could be repeated throughout the brigade
     * (coalesce any run of consecutive data buckets) but this would
     * add significant complexity, particularly to memory
     * management. */
    for (e = APR_BRIGADE_FIRST(bb);
         e != APR_BRIGADE_SENTINEL(bb)
             && !APR_BUCKET_IS_METADATA(e)
             && e->length != (apr_size_t)-1
             && e->length <= COALESCE_BYTES
             && (buffered + bytes + e->length) <= COALESCE_BYTES;
         e = APR_BUCKET_NEXT(e)) {
        /* don't count zero-length buckets */
        if (e->length) {
            bytes += e->length;
            count++;
        }
    }

    /* If there is room remaining and the next bucket is a data
     * bucket, try to include it in the prefix to coalesce.  For a
     * typical [HEAP] [FILE] HTTP response brigade, this handles
     * merging the headers and the start of the body into a single TLS
     * record. */
    if (bytes + buffered > 0
        && bytes + buffered < COALESCE_BYTES
        && e != APR_BRIGADE_SENTINEL(bb)
        && !APR_BUCKET_IS_METADATA(e)) {
        apr_status_t rv = APR_SUCCESS;

        /* For an indeterminate length bucket (PIPE/CGI/...), try a
         * non-blocking read to have it morph into a HEAP.  If the
         * read fails with EAGAIN, it is harmless to try a split
         * anyway, split is ENOTIMPL for most PIPE-like buckets. */
        if (e->length == (apr_size_t)-1) {
            const char *discard;
            apr_size_t ignore;

            rv = apr_bucket_read(e, &discard, &ignore, APR_NONBLOCK_READ);
            if (rv != APR_SUCCESS && !APR_STATUS_IS_EAGAIN(rv)) {
                ap_log_cerror(APLOG_MARK, APLOG_ERR, rv, f->c, APLOGNO(10232)
                              "coalesce failed to read from %s bucket",
                              e->type->name);
                return AP_FILTER_ERROR;
            }
        }

        if (rv == APR_SUCCESS) {
            /* If the read above made the bucket morph, it may now fit
             * entirely within the buffer.  Otherwise, split it so it does
             * fit. */
            if (e->length > COALESCE_BYTES
                || e->length + buffered + bytes > COALESCE_BYTES) {
                rv = apr_bucket_split(e, COALESCE_BYTES - (buffered + bytes));
            }

            if (rv == APR_SUCCESS && e->length == 0) {
                /* As above, don't count in the prefix if the bucket is
                 * now zero-length. */
            }
            else if (rv == APR_SUCCESS) {
                ap_log_cerror(APLOG_MARK, APLOG_TRACE4, 0, f->c,
                              "coalesce: adding %" APR_SIZE_T_FMT " bytes "
                              "from split %s bucket, total %" APR_SIZE_T_FMT,
                              e->length, e->type->name, bytes + buffered);

                count++;
                bytes += e->length;
                e = APR_BUCKET_NEXT(e);
            }
--> --------------------

--> maximum size reached

--> --------------------

quality98%

¤ Dauer der Verarbeitung: 0.41 Sekunden (vorverarbeitet) ¤

Wurzel

Suchen

Beweissystem der NASA

Beweissystem Isabelle

NIST Cobol Testsuite

Cephes Mathematical Library

Wiener Entwicklungsmethode

Haftungshinweis

Die Informationen auf dieser Webseite wurden nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit, noch Qualität der bereit gestellten Informationen zugesichert.

Bemerkung:

Die farbliche Syntaxdarstellung ist noch experimentell.