blob: dc2bba962bb2e4912ff6d72ae331cf25507f2173 [file] [log] [blame]
/* 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.
*/
#include "apr.h"
#include "apr_strings.h"
#include "apr_buckets.h"
#include "apr_lib.h"
#include "apr_poll.h"
#define APR_WANT_STRFUNC
#define APR_WANT_MEMFUNC
#include "apr_want.h"
#include "httpd.h"
#include "util_filter.h"
/* Read and discard all output from the brigade. Note that with the
* CGI bucket, the brigade will become empty once the script's stdout
* is closed (or on error/timeout), but the stderr output may not have
* been entirely captured at this point. */
static void discard_script_output(apr_bucket_brigade *bb)
{
apr_bucket *e;
const char *buf;
apr_size_t len;
for (e = APR_BRIGADE_FIRST(bb);
e != APR_BRIGADE_SENTINEL(bb) && !APR_BUCKET_IS_EOS(e);
e = APR_BRIGADE_FIRST(bb))
{
if (apr_bucket_read(e, &buf, &len, APR_BLOCK_READ)) {
break;
}
apr_bucket_delete(e);
}
}
#ifdef WANT_CGI_BUCKET
/* A CGI bucket type is needed to catch any output to stderr from the
* script; see PR 22030. */
static const apr_bucket_type_t bucket_type_cgi;
struct cgi_bucket_data {
apr_pollset_t *pollset;
request_rec *r;
apr_interval_time_t timeout;
};
/* Create a CGI bucket using pipes from script stdout 'out'
* and stderr 'err', for request 'r'. */
static apr_bucket *cgi_bucket_create(request_rec *r,
apr_interval_time_t timeout,
apr_file_t *out, apr_file_t *err,
apr_bucket_alloc_t *list)
{
apr_bucket *b = apr_bucket_alloc(sizeof(*b), list);
apr_status_t rv;
apr_pollfd_t fd;
struct cgi_bucket_data *data = apr_palloc(r->pool, sizeof *data);
/* Disable APR timeout handling since we'll use poll() entirely. */
apr_file_pipe_timeout_set(out, 0);
apr_file_pipe_timeout_set(err, 0);
APR_BUCKET_INIT(b);
b->free = apr_bucket_free;
b->list = list;
b->type = &bucket_type_cgi;
b->length = (apr_size_t)(-1);
b->start = -1;
/* Create the pollset */
rv = apr_pollset_create(&data->pollset, 2, r->pool, 0);
if (rv != APR_SUCCESS) {
ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r, APLOGNO(01217)
"apr_pollset_create(); check system or user limits");
return NULL;
}
fd.desc_type = APR_POLL_FILE;
fd.reqevents = APR_POLLIN;
fd.p = r->pool;
fd.desc.f = out; /* script's stdout */
fd.client_data = (void *)1;
rv = apr_pollset_add(data->pollset, &fd);
if (rv != APR_SUCCESS) {
ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r, APLOGNO(01218)
"apr_pollset_add(); check system or user limits");
return NULL;
}
fd.desc.f = err; /* script's stderr */
fd.client_data = (void *)2;
rv = apr_pollset_add(data->pollset, &fd);
if (rv != APR_SUCCESS) {
ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r, APLOGNO(01219)
"apr_pollset_add(); check system or user limits");
return NULL;
}
data->r = r;
data->timeout = timeout;
b->data = data;
return b;
}
/* Create a duplicate CGI bucket using given bucket data */
static apr_bucket *cgi_bucket_dup(struct cgi_bucket_data *data,
apr_bucket_alloc_t *list)
{
apr_bucket *b = apr_bucket_alloc(sizeof(*b), list);
APR_BUCKET_INIT(b);
b->free = apr_bucket_free;
b->list = list;
b->type = &bucket_type_cgi;
b->length = (apr_size_t)(-1);
b->start = -1;
b->data = data;
return b;
}
/* Handle stdout from CGI child. Duplicate of logic from the _read
* method of the real APR pipe bucket implementation. */
static apr_status_t cgi_read_stdout(apr_bucket *a, apr_file_t *out,
const char **str, apr_size_t *len)
{
char *buf;
apr_status_t rv;
*str = NULL;
*len = APR_BUCKET_BUFF_SIZE;
buf = apr_bucket_alloc(*len, a->list); /* XXX: check for failure? */
rv = apr_file_read(out, buf, len);
if (rv != APR_SUCCESS && rv != APR_EOF) {
apr_bucket_free(buf);
return rv;
}
if (*len > 0) {
struct cgi_bucket_data *data = a->data;
apr_bucket_heap *h;
/* Change the current bucket to refer to what we read */
a = apr_bucket_heap_make(a, buf, *len, apr_bucket_free);
h = a->data;
h->alloc_len = APR_BUCKET_BUFF_SIZE; /* note the real buffer size */
*str = buf;
APR_BUCKET_INSERT_AFTER(a, cgi_bucket_dup(data, a->list));
}
else {
apr_bucket_free(buf);
a = apr_bucket_immortal_make(a, "", 0);
*str = a->data;
}
return rv;
}
/* Read method of CGI bucket: polls on stderr and stdout of the child,
* sending any stderr output immediately away to the error log. */
static apr_status_t cgi_bucket_read(apr_bucket *b, const char **str,
apr_size_t *len, apr_read_type_e block)
{
struct cgi_bucket_data *data = b->data;
apr_interval_time_t timeout = 0;
apr_status_t rv;
int gotdata = 0;
if (block != APR_NONBLOCK_READ) {
timeout = data->timeout > 0 ? data->timeout : data->r->server->timeout;
}
do {
const apr_pollfd_t *results;
apr_int32_t num;
rv = apr_pollset_poll(data->pollset, timeout, &num, &results);
if (APR_STATUS_IS_TIMEUP(rv)) {
if (timeout) {
ap_log_rerror(APLOG_MARK, APLOG_WARNING, 0, data->r, APLOGNO(01220)
"Timeout waiting for output from CGI script %s",
data->r->filename);
return rv;
}
else {
return APR_EAGAIN;
}
}
else if (APR_STATUS_IS_EINTR(rv)) {
continue;
}
else if (rv != APR_SUCCESS) {
ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, data->r, APLOGNO(01221)
"poll failed waiting for CGI child");
return rv;
}
for (; num; num--, results++) {
if (results[0].client_data == (void *)1) {
/* stdout */
rv = cgi_read_stdout(b, results[0].desc.f, str, len);
if (APR_STATUS_IS_EOF(rv)) {
rv = APR_SUCCESS;
}
gotdata = 1;
} else {
/* stderr */
apr_status_t rv2 = log_script_err(data->r, results[0].desc.f);
if (APR_STATUS_IS_EOF(rv2)) {
apr_pollset_remove(data->pollset, &results[0]);
}
}
}
} while (!gotdata);
return rv;
}
static const apr_bucket_type_t bucket_type_cgi = {
"CGI", 5, APR_BUCKET_DATA,
apr_bucket_destroy_noop,
cgi_bucket_read,
apr_bucket_setaside_notimpl,
apr_bucket_split_notimpl,
apr_bucket_copy_notimpl
};
#endif /* WANT_CGI_BUCKET */
/* Handle the CGI response output, having set up the brigade with the
* CGI or PIPE bucket as appropriate. */
static int cgi_handle_response(request_rec *r, int nph, apr_bucket_brigade *bb,
apr_interval_time_t timeout, cgi_server_conf *conf,
char *logdata, apr_file_t *script_err)
{
apr_status_t rv;
/* Handle script return... */
if (!nph) {
const char *location;
char sbuf[MAX_STRING_LEN];
int ret;
if ((ret = ap_scan_script_header_err_brigade_ex(r, bb, sbuf,
APLOG_MODULE_INDEX)))
{
/* In the case of a timeout reading script output, clear
* the brigade to avoid a second attempt to read the
* output. */
if (ret == HTTP_GATEWAY_TIME_OUT) {
apr_brigade_cleanup(bb);
}
ret = log_script(r, conf, ret, logdata, sbuf, bb, script_err);
/*
* ret could be HTTP_NOT_MODIFIED in the case that the CGI script
* does not set an explicit status and ap_meets_conditions, which
* is called by ap_scan_script_header_err_brigade, detects that
* the conditions of the requests are met and the response is
* not modified.
* In this case set r->status and return OK in order to prevent
* running through the error processing stack as this would
* break with mod_cache, if the conditions had been set by
* mod_cache itself to validate a stale entity.
* BTW: We circumvent the error processing stack anyway if the
* CGI script set an explicit status code (whatever it is) and
* the only possible values for ret here are:
*
* HTTP_NOT_MODIFIED (set by ap_meets_conditions)
* HTTP_PRECONDITION_FAILED (set by ap_meets_conditions)
* HTTP_INTERNAL_SERVER_ERROR (if something went wrong during the
* processing of the response of the CGI script, e.g broken headers
* or a crashed CGI process).
*/
if (ret == HTTP_NOT_MODIFIED) {
r->status = ret;
return OK;
}
return ret;
}
location = apr_table_get(r->headers_out, "Location");
if (location && r->status == 200) {
/* For a redirect whether internal or not, discard any
* remaining stdout from the script, and log any remaining
* stderr output, as normal. */
discard_script_output(bb);
apr_brigade_destroy(bb);
if (script_err) {
apr_file_pipe_timeout_set(script_err, timeout);
log_script_err(r, script_err);
}
}
if (location && location[0] == '/' && r->status == 200) {
/* This redirect needs to be a GET no matter what the original
* method was.
*/
r->method = "GET";
r->method_number = M_GET;
/* We already read the message body (if any), so don't allow
* the redirected request to think it has one. We can ignore
* Transfer-Encoding, since we used REQUEST_CHUNKED_ERROR.
*/
apr_table_unset(r->headers_in, "Content-Length");
ap_internal_redirect_handler(location, r);
return OK;
}
else if (location && r->status == 200) {
/* XXX: Note that if a script wants to produce its own Redirect
* body, it now has to explicitly *say* "Status: 302"
*/
discard_script_output(bb);
apr_brigade_destroy(bb);
return HTTP_MOVED_TEMPORARILY;
}
rv = ap_pass_brigade(r->output_filters, bb);
}
else /* nph */ {
struct ap_filter_t *cur;
/* get rid of all filters up through protocol... since we
* haven't parsed off the headers, there is no way they can
* work
*/
cur = r->proto_output_filters;
while (cur && cur->frec->ftype < AP_FTYPE_CONNECTION) {
cur = cur->next;
}
r->output_filters = r->proto_output_filters = cur;
rv = ap_pass_brigade(r->output_filters, bb);
}
/* don't soak up script output if errors occurred writing it
* out... otherwise, we prolong the life of the script when the
* connection drops or we stopped sending output for some other
* reason */
if (script_err && rv == APR_SUCCESS && !r->connection->aborted) {
apr_file_pipe_timeout_set(script_err, timeout);
log_script_err(r, script_err);
}
if (script_err) apr_file_close(script_err);
return OK; /* NOT r->status, even if it has changed. */
}
/* Read the request body and write it to fd 'script_out', using 'bb'
* as temporary bucket brigade. If 'logbuf' is non-NULL, the first
* logbufbytes of stdout are stored in logbuf. */
static apr_status_t cgi_handle_request(request_rec *r, apr_file_t *script_out,
apr_bucket_brigade *bb,
char *logbuf, apr_size_t logbufbytes)
{
int seen_eos = 0;
int child_stopped_reading = 0;
apr_status_t rv;
int dbpos = 0;
do {
apr_bucket *bucket;
rv = ap_get_brigade(r->input_filters, bb, AP_MODE_READBYTES,
APR_BLOCK_READ, HUGE_STRING_LEN);
if (rv != APR_SUCCESS) {
return rv;
}
for (bucket = APR_BRIGADE_FIRST(bb);
bucket != APR_BRIGADE_SENTINEL(bb);
bucket = APR_BUCKET_NEXT(bucket))
{
const char *data;
apr_size_t len;
if (APR_BUCKET_IS_EOS(bucket)) {
seen_eos = 1;
break;
}
/* We can't do much with this. */
if (APR_BUCKET_IS_FLUSH(bucket)) {
continue;
}
/* If the child stopped, we still must read to EOS. */
if (child_stopped_reading) {
continue;
}
/* read */
rv = apr_bucket_read(bucket, &data, &len, APR_BLOCK_READ);
if (rv) {
return rv;
}
if (logbufbytes && dbpos < logbufbytes) {
int cursize;
if ((dbpos + len) > logbufbytes) {
cursize = logbufbytes - dbpos;
}
else {
cursize = len;
}
memcpy(logbuf + dbpos, data, cursize);
dbpos += cursize;
}
/* Keep writing data to the child until done or too much time
* elapses with no progress or an error occurs.
*/
rv = apr_file_write_full(script_out, data, len, NULL);
if (rv != APR_SUCCESS) {
/* silly script stopped reading, soak up remaining message */
child_stopped_reading = 1;
ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r, APLOGNO(02651)
"Error writing request body to script %s",
r->filename);
}
}
apr_brigade_cleanup(bb);
}
while (!seen_eos);
if (logbuf) {
logbuf[dbpos] = '\0';
}
return APR_SUCCESS;
}