Fixed potential segfault in ngx_http_keepalive_handler().
In case of error in the read event handling we close a connection
by calling ngx_http_close_connection(), that also destroys connection
pool. Thereafter, an attempt to free a buffer (added in r4892) that
was allocated from the pool could cause SIGSEGV and is meaningless
as well (the buffer already freed with the pool).
SSL: retry "sess_id" and "id" allocations.
In case of fully populated SSL session cache with no memory left for
new allocations, ngx_ssl_new_session() will try to expire the oldest
non-expired session and retry, but only in case when slab allocation
fails for "cached_sess", not when slab allocation fails for either
"sess_id" or "id", which can happen for number of reasons and results
in new session not being cached.
Patch fixes this by adding retry logic to "sess_id" & "id" allocations.
Patch by Piotr Sikora.
It was added in r2717 and no longer needed since r2721,
where the termination was added to ngx_shm_alloc() and
ngx_init_zone_pool(). Since then it only corrupted error
messages about invalid zones.
Configure: rebuild perl module nginx.so if headers are changed.
Note: the "-p" argument of cp(1) dropped intentionally, to force nginx.so
rebuild. It is considered too boring to properly list all dependencies
in Makefile.PL.
*) FastCGI: fixed wrong connection close with fastcgi_keep_conn.
With fastcgi_keep_conn it was possible that connection was closed after
FCGI_STDERR record with zero padding and without any further data read
yet. This happended as f->state was set to ngx_http_fastcgi_st_padding
and then "break" happened, resulting in p->length being set to
f->padding, i.e. 0 (which in turn resulted in connection close).
Fix is to make sure we continue the loop after f->state is set.
*) FastCGI: unconditional state transitions. Checks for f->padding
before state transitions make code hard to follow, remove them and
make sure we always do another loop iteration after f->state is
set to ngx_http_fastcgi_st_padding.
*) FastCGI: proper handling of split fastcgi end request. If fastcgi
end request record was split between several network packets, with
fastcgi_keep_conn it was possible that connection was saved in
incorrect state (e.g. with padding bytes not yet read).
The "secure_link_secret" directive was always inherited from the outer
configuration level even when "secure_link" and "secure_link_md5" were
specified on the inner level.
Events: fixed null pointer dereference with resolver and poll.
A POLLERR signalled by poll() without POLLIN/POLLOUT, as seen on
Linux, would generate both read and write events, but there's no
write event handler for resolver events. A fix is to only call
event handler of an active event.
*) Configure: fixed style of include directories.
*) Configure: fixed GeoIP library detection.
*) GeoIP: IPv6 support. When using IPv6 databases, IPv4 addresses are
looked up as IPv4-mapped IPv6 addresses. Mostly based on a patch
by Gregor Kališnik (ticket #250).
*) GeoIP: removed pseudo-support of "proxy" and "netspeed" databases.
Before the patch if proxy_method was specified at http{} level the code
to add trailing space wasn't executed, resulting in incorrect requests
to upstream.
*) SSL: speedup loading of configs with many ssl servers. The patch
saves one EC_KEY_generate_key() call per server{} block by informing
OpenSSL about SSL_OP_SINGLE_ECDH_USE we are going to use before
the SSL_CTX_set_tmp_ecdh() call.
For a configuration file with 10k simple server{} blocks with SSL
enabled this change reduces startup time from 18s to 5s on a slow
test box here.
*) SSL: removed conditions that always hold true.
*) SSL: resetting of flush flag after the data was written. There is
no need to flush next chunk of data if it does not contain a buffer
with the flush or last_buf flags set.
*) SSL: preservation of flush flag for buffered data. Previously,
if SSL buffer was not sent we lost information that the data
must be flushed.
*) SSL: calculation of buffer size moved closer to its usage.
No functional changes.
*) SSL: avoid calling SSL_write() with zero data size. According to
documentation, calling SSL_write() with num=0 bytes to be sent
results in undefined behavior.
We don't currently call ngx_ssl_send_chain() with empty chain and
buffer. This check handles the case of a chain with total data size
that is a multiple of NGX_SSL_BUFSIZE, and with the special buffer
at the end.
In practice such cases resulted in premature connection close and
critical error "SSL_write() failed (SSL:)" in the error log.
*) SSL: take into account data in the buffer while limiting output.
In some rare cases this can result in a more smooth sending rate.
*) SSL: fixed ngx_ssl_handshake() with level-triggered event methods.
Missing calls to ngx_handle_write_event() and ngx_handle_read_event()
resulted in a CPU hog during SSL handshake if an level-triggered event
method (e.g. select) was used.
Fixed off-by-one during upstream state resetting when switching to
backup servers if there were exactly 32 (64 on 64-bit platforms)
backup servers configured.
Based on patch by Thomas Chen (ticket #257).
*) Access log: fixed redundant buffer reallocation. Previously a new
buffer was allocated for every "access_log" directive with the same
file path and "buffer=" parameters, while only one buffer per file
is used.
*) Reopening log files code moved to a separate function. The code
refactored in a way to call custom handler that can do appropriate
cleanup work (if any), like flushing buffers, finishing compress
streams, finalizing connections to log daemon, etc..
*) Access log: the "flush" parameter of the "access_log" directive.
*) Configure: added the NGX_ZLIB define. This was introduced for
conditional compilation of the code that requires the zlib library.
*) Access log: the "gzip" parameter of the "access_log" directive.
Note: this requires zlib version 1.2.0.4 or above to work.
*) The data pointer in ngx_open_file_t objects must be initialized.
Uninitialized pointer may result in arbitrary segfaults if access_log
is used without buffer and without variables in file path.
Patch by Tatsuhiko Kubo (ticket #268).
*) Image filter: configuration inheritance fixes.
The image_filter_jpeg_quality, image_filter_sharpen and
"image_filter rotate" were inherited incorrectly if a directive
with variables was defined, and then redefined to a literal value,
i.e. in configurations like
image_filter_jpeg_quality $arg_q;
location / {
image_filter_jpeg_quality 50;
}
Patch by Ian Babrou, with minor changes.
*) Image filter: fixed image_filter rotate inheritance.
Configurations like
location /i/ {
image_filter resize 200 200;
image_filter rotate 180;
location /i/foo/ {
image_filter resize 200 200;
}
}
resulted in rotation incorrectly applied in the location /i/foo,
without any way to clear it. Fix is to handle conf->angle/conf->acv
consistently with other filter variables and do not try to inherit
them if there are transformations defined for current location.
The crypt_r() function returns NULL on errors, check it explicitly instead
of assuming errno will remain 0 if there are no errors (per POSIX, the
setting of errno after a successful call to a function is unspecified
unless the description of that function specifies that errno shall not
be modified).
Additionally, dropped unneeded ngx_set_errno(0) and fixed error handling
of memory allocation after normal crypt(), which was inapropriate and
resulted in null pointer dereference on allocation failures.
*) Geo: IPv6 support. The "ranges" mode is still limited to IPv4 only.
*) Geo: properly initialize ngx_cidr_t when dealing with "default".
*) Geo: made "default" affect both IPv4 and IPv6 when using prefixes.
Previously, "default" was equivalent to specifying 0.0.0.0/0, now
it's equivalent to specifying both 0.0.0.0/0 and ::/0 (if support
for IPv6 is enabled) with the same value.
*) Geo: improved code readability.
*) Geo: improved ngx_http_geo_block() code readability.
*) Geo: fixed the "ranges" without ranges case. The following
configuration returned an empty value for $geo:
geo $geo {
ranges;
default default;
}
*) Fixed return type of internal function that allocates radix tree nodes.
*) There's no need to normalize address returned by ngx_ptocidr().
*) Geo: ensure that default entry is always present. If 0.0.0.0/32 entry
was present and there was no explicit "default", we failed to add an
empty string as a default value.
*) Trailing whitespace fix.
The ngx_write_fd() and ngx_read_fd() functions return -1 in case of error,
so the incorrect comparison with NGX_FILE_ERROR (which is 0 on windows
platforms) might result in inaccurate error message in the error log.
Also the ngx_errno global variable is being set only if the returned value
is -1.
An incorrect memLevel (lower than 1) might be passed to deflateInit2() if the
"gzip_hash" directive is set to a value less than the value of "gzip_window"
directive. This resulted in "deflateInit2() failed: -2" alert and an empty
reply.
Configuration like
location / {
set $true 1;
if ($true) {
proxy_pass http://backend;
}
if ($true) {
# nothing
}
}
resulted in segmentation fault due to NULL pointer dereference as the
upstream configuration wasn't initialized in an implicit location created
by the last if(), but the r->content_handler was set due to first if().
Instead of committing a suicide by dereferencing a NULL pointer, return
500 (Internal Server Error) in such cases, i.e. if uscf is NULL. Better
fix would be to avoid such cases by fixing the "if" directive handling,
but it's out of scope of this patch.
Prodded by Piotr Sikora.
On Mac OS X system toolchain by default prefers include files
from /usr/local/include, but libraries from /usr/lib. This might result in
various problems, in particular the one outlined below.
If the PCRE library is installed into /usr/local/, this results in pcre.h
being used from /usr/local/include (with PCRE_CONFIG_JIT defined), but
libpcre from /usr/lib (as shipped with the OS, without pcre_free_study()
symbol). As a result build fails as we use pcre_free_study() function
if we try to compile with PCRE JIT support.
Obvious workaround to the root cause is to ask compiler to prefer
library from /usr/local/lib via ./configure --with-ld-opt="-L/usr/local/lib".
On the other hand, in any case it would be good to check if the function
we are going to use is available, hence the change.
See thread here for details:
http://mailman.nginx.org/pipermail/nginx-devel/2012-December/003074.html
Prodded by Piotr Sikora.
*) Allow the complex value to be defined as an empty string.
This makes conversion from strings to complex values possible
without the loss of functionality.
*) The "auth_basic" directive gained support of variables.
*) Fixed variable syntax checking in "set", "geo", "limit_conn_zone",
and "perl_set" directives.
*) Added checks that disallow adding a variable with an empty name.
Added variable name syntax checks to "geo" and "map" directives.
*) Variables $pipe, $request_length, $time_iso8601, and $time_local.
Log module counterparts are preserved for efficiency.
Based on patch by Kiril Kalchev.
*) Fixed the NGX_SOCKADDR_STRLEN macro definition.
The ngx_sock_ntop() function, when told to print both address and port,
prints IPv6 address in square brackets, followed by colon and port.
*) Properly initialize "struct in6_addr" with zeroes.
This will result in alphabetical sorting of included files if
the "include" directive with wildcards is used.
Note that the behaviour is now different from that on Windows, where
alphabetical sorting is not guaranteed for FindFirsFile()/FindNextFile()
(used to be alphabetical on NTFS, but not on FAT).
Approved by Igor Sysoev, prodded by many.
*) Request body: fixed "501 Not Implemented" error handling.
It is not about "Method" but a generic message, and is expected to be used
e.g. if specified Transfer-Encoding is not supported. Fixed message to
match RFC 2616.
Additionally, disable keepalive on such errors as we won't be able to read
request body correctly if we don't understand Transfer-Encoding used.
*) Request body: $request_body variable generalization.
The $request_body variable was assuming there can't be more than two
buffers. While this is currently true due to request body reading
implementation details, this is not a good thing to depend on and may
change in the future.
*) Request body: code duplication reduced, no functional changes.
The r->request_body_in_file_only with empty body case is now handled in
ngx_http_write_request_body().
*) Request body: fixed socket leak on errors.
The r->main->count reference counter was always incremented in
ngx_http_read_client_request_body(), while it is only needs to be
incremented on positive returns.
*) Request body: properly handle events while discarding body.
An attempt to call ngx_handle_read_event() before actually reading
data from a socket might result in read event being disabled, which is
wrong. Catched by body.t test on Solaris.
*) Fixed location of debug message in ngx_shmtx_lock().
*) Core: don't reuse shared memory zone that changed ownership (ticket #210).
nginx doesn't allow the same shared memory zone to be used for different
purposes, but failed to check this on reconfiguration. If a shared memory
zone was used for another purpose in the new configuration, nginx attempted
to reuse it and crashed.
*) Core: added debug logging of writev() in ngx_write_chain_to_file().
*) Core: fixed ngx_write_chain_to_file() with IOV_MAX reached.
Catched by dav_chunked.t on Solaris. In released versions this might
potentially result in corruption of complex protocol responses if they
were written to disk and there were more distinct buffers than IOV_MAX
in a single write.
Dav: fixed segfault on PUT if body was already read (ticket #238).
If request body reading happens with different options it's possible
that there will be no r->request_body->temp_file available (or even
no r->request_body available if body was discarded). Return internal
server error in this case instead of committing suicide by dereferencing
a null pointer.
*) Upstream: honor the "down" flag for a single server.
If an upstream block was defined with the only server marked as
"down", e.g.
upstream u {
server 127.0.0.1:8080 down;
}
an attempt was made to contact the server despite the "down" flag.
It is believed that immediate 502 response is better in such a
case, and it's also consistent with what is currently done in case
of multiple servers all marked as "down".
*) Upstream: better detection of connect() failures with kqueue.
Pending EOF might be reported on both read and write events, whichever
comes first, so check both of them.
Patch by Yichun Zhang (agentzh), slightly modified.
