Andy Green [Wed, 12 Oct 2016 22:32:57 +0000 (06:32 +0800)]
Add reject service keywords list
Andy Green [Mon, 10 Oct 2016 12:34:34 +0000 (20:34 +0800)]
adjust_timeout: with default lws_plat_service_tsi allow beings passed 0 timeout
Some people are calling service with zero timeout, taking care of
not busywaiting by some other external arrangements.
Adapt the forced service signalling to survive this.
Joerg Pommnitz [Mon, 10 Oct 2016 12:10:39 +0000 (20:10 +0800)]
some compilers need void param explicitly
Yuchen Xie [Mon, 10 Oct 2016 12:10:34 +0000 (20:10 +0800)]
Correct the library name of LIBHUBBUB_LIBRARIES
It should be `hubbub` in `find_library` to make the function work.
Andy Green [Sat, 8 Oct 2016 10:08:03 +0000 (18:08 +0800)]
docs: explain lws_write handling of truncated sends better
Denis Osvald [Fri, 7 Oct 2016 09:27:46 +0000 (11:27 +0200)]
publicly document lws_service_fd timeout servicing
Signed-off-by: Denis Osvald <denis.osvald@sartura.hr>
Denis Osvald [Thu, 6 Oct 2016 13:45:25 +0000 (15:45 +0200)]
test-server-extpoll: add 1-per-second timeout servicing
Signed-off-by: Denis Osvald <denis.osvald@sartura.hr>
Denis Osvald [Fri, 7 Oct 2016 08:57:02 +0000 (10:57 +0200)]
test-server-extpoll: check for forced service before looping again
Signed-off-by: Denis Osvald <denis.osvald@sartura.hr>
Andy Green [Thu, 6 Oct 2016 13:48:20 +0000 (21:48 +0800)]
v2.1.0
Bump soname to 9
Andy Green [Thu, 6 Oct 2016 19:19:17 +0000 (03:19 +0800)]
extpoll: expose forced service apis
Andy Green [Thu, 6 Oct 2016 12:40:28 +0000 (20:40 +0800)]
update attack.sh
Lws cares about trailing \n on a lot of these tests now. Make it check it still cares on one and remove
the trailing \n on the others.
There's 2 changes in the results about /..//?, it seems to apply / to uri arg 1. But it doesn't seem
to make a problem so just adapt the results for now.
Andy Green [Wed, 5 Oct 2016 05:49:46 +0000 (13:49 +0800)]
ubuntu: fix uv detection for lwsws
Andy Green [Wed, 5 Oct 2016 01:52:39 +0000 (09:52 +0800)]
travis: explicitly point to openssl on osx
One day this started failing at CMake autofind. This forces it to look at the right place.
Andy Green [Wed, 5 Oct 2016 02:15:16 +0000 (10:15 +0800)]
fix cosmetic if end if name tag error
Andy Green [Tue, 4 Oct 2016 10:05:10 +0000 (18:05 +0800)]
closing drops any pending ah rx immediately
Andy Green [Tue, 4 Oct 2016 00:39:14 +0000 (08:39 +0800)]
port forced service checking from libuv
Related to second part of
https://github.com/warmcat/libwebsockets/issues/638
Andy Green [Tue, 4 Oct 2016 00:24:00 +0000 (08:24 +0800)]
post form parsing fix retry as new boundary start needed after mismatching boundary
https://github.com/warmcat/libwebsockets/issues/641
Peter Pentchev [Sat, 1 Oct 2016 23:21:20 +0000 (02:21 +0300)]
Remove the cleanup functions with OpenSSL 1.1.
The thread support in OpenSSL has been rewritten almost completely
and the cleanup functions are now executed automatically.
Peter Pentchev [Sat, 1 Oct 2016 23:21:03 +0000 (02:21 +0300)]
Subject: Fix some typographical and grammatical errors.
Andy Green [Thu, 29 Sep 2016 02:31:06 +0000 (10:31 +0800)]
lws_header_table_reset: make caller responsibility to clear down ah rx buffer
There are two kinds of reaason to call lws_header_table_reset(), one is we are reallocating
a destroyed ah to another wsi, and the other is we are moving to the next pipelined header set
still on the same wsi, and we need a "weaker" reset that only clears down the state related
to the header parsing, not everything about the ah context including the ah rx buffer.
This patch moves the ah rxbuffer rxpos and rxlen resetting out of lws_header_table_reset() and to
be the responsibility of the caller. Callers who are moving the ah to another wsi are
patched to deal with resetting rxpos and rxlen and lws_http_transaction_completed() who only
resets the ah when moving to the next pipelined headers, no longer wrongly clears the ah rxbuf.
https://github.com/warmcat/libwebsockets/issues/638
Brown, Matthew [Mon, 26 Sep 2016 21:32:40 +0000 (05:32 +0800)]
Added option to build the static library with PIC
Benjamin Ness [Mon, 26 Sep 2016 19:18:16 +0000 (14:18 -0500)]
fix build problem on systems without X509_VERIFY_PARAM type
Patrick Gansterer [Thu, 22 Sep 2016 22:04:40 +0000 (00:04 +0200)]
Add SVG to lws_get_mimetype()
Andy Green [Mon, 19 Sep 2016 11:16:47 +0000 (19:16 +0800)]
lwsws license to cc0
https://github.com/warmcat/libwebsockets/issues/629
Andy Green [Wed, 14 Sep 2016 18:22:57 +0000 (02:22 +0800)]
lws_snprintf
Thanks to Fabrice Gilot for reporting the problem that led to uncovering this.
Due to a misunderstanding of the return value of snprintf (it is not truncated according
to the max size passed in) in several places relying on snprintf to truncate the length
overflows are possible.
This patch wraps snprintf with a new lws_snprintf() which does truncate its length to allow
the buffer limiting scheme to work properly.
All users should update with these fixes.
Andy Green [Fri, 9 Sep 2016 20:43:07 +0000 (04:43 +0800)]
handle rx flow control active when consuming payload
https://github.com/warmcat/libwebsockets/issues/622
Andy Green [Fri, 9 Sep 2016 20:27:27 +0000 (04:27 +0800)]
generic-table: format-security
https://github.com/warmcat/libwebsockets/issues/624
Andy Green [Fri, 9 Sep 2016 14:15:43 +0000 (22:15 +0800)]
lws_status protect against pss list changing
Andy Green [Fri, 9 Sep 2016 00:29:32 +0000 (08:29 +0800)]
adopt_socket_vhost: error path doesn't remove us from timeout list
As found by "github user 7"
https://github.com/warmcat/libwebsockets/issues/621
Fredrik Skogman [Thu, 8 Sep 2016 22:51:43 +0000 (06:51 +0800)]
Do not use ps fax on Solaris.
Fredrik Skogman [Thu, 8 Sep 2016 22:50:41 +0000 (06:50 +0800)]
Updated test programs to build on Solaris. Some whitespaces cleanup.
Fredrik Skogman [Thu, 8 Sep 2016 22:49:44 +0000 (06:49 +0800)]
Added build support for Solaris.
Fredrik Skogman [Thu, 8 Sep 2016 22:48:24 +0000 (06:48 +0800)]
Check for sys/sockio.h.
Andy Green [Tue, 6 Sep 2016 07:36:51 +0000 (15:36 +0800)]
client http: extra read notification after close
https://github.com/warmcat/libwebsockets/issues/620
Andy Green [Mon, 5 Sep 2016 07:03:37 +0000 (15:03 +0800)]
windows: WCHAR in lws_plat_inet_ntop needs double the final allocation
https://github.com/warmcat/libwebsockets/issues/619
Andy Green [Sun, 28 Aug 2016 01:39:21 +0000 (09:39 +0800)]
coverity 169276-9 - false positive assuming 8b char: char limits index size
Andy Green [Sun, 28 Aug 2016 01:34:27 +0000 (09:34 +0800)]
coverity 169274 - lwsgt dirlisting ignore files that cant be statted
Andy Green [Sun, 28 Aug 2016 01:28:55 +0000 (09:28 +0800)]
coverity 169273 - off-by-one on cgi chunking swallow limit
Andy Green [Sun, 28 Aug 2016 01:24:06 +0000 (09:24 +0800)]
coverity 169272 - off-by-one possible on CGI buffer limit
Andy Green [Sun, 28 Aug 2016 01:21:56 +0000 (09:21 +0800)]
coverity 169271 - take care about sockfd of -1 on close
Andy Green [Sun, 28 Aug 2016 01:18:17 +0000 (09:18 +0800)]
coverity 169269 - dont issue NULL cce
Andy Green [Sun, 28 Aug 2016 01:15:26 +0000 (09:15 +0800)]
coverity 169268 + 169270- dead code plus repeat NULL check on error path
Andy Green [Sun, 28 Aug 2016 01:12:39 +0000 (09:12 +0800)]
coverity 169275 - lwsgt check string bounds
Andy Green [Sun, 28 Aug 2016 01:10:40 +0000 (09:10 +0800)]
coverity 169267 - lwsgt strncpy may fill buffer without NUL
Andy Green [Sat, 27 Aug 2016 09:07:06 +0000 (17:07 +0800)]
per-vhost headers and lwsws conf support
This l;ets you add per-vhost arbitrary headers on served files, eg
"headers": [{
"X-Content-Type-Options": "nosniff",
"X-XSS-Protection": "1; mode=block",
"x-frame-options": "SAMEORIGIN"
}],
Patrick Gansterer [Fri, 26 Aug 2016 21:42:04 +0000 (23:42 +0200)]
Update badges in README.md
Fix the link to Appveyor and use SVG images.
Patrick Gansterer [Fri, 26 Aug 2016 08:49:29 +0000 (10:49 +0200)]
Remove unneeded #include <stdint.h>
This fixes the build for Visual Studio 2008.
Andy Green [Tue, 23 Aug 2016 10:29:17 +0000 (18:29 +0800)]
test html: js date uses ms
Andy Green [Tue, 23 Aug 2016 06:20:11 +0000 (14:20 +0800)]
cgi-retain-timeout-after-POST-send
Sending the POST data isn't enough to let us off the hook for timeout checks, if we
are doing a CGI on it.
Andy Green [Sun, 21 Aug 2016 23:07:10 +0000 (07:07 +0800)]
quench logging
Andy Green [Fri, 19 Aug 2016 22:34:46 +0000 (06:34 +0800)]
cgi: deal with all methods correctly
Andy Green [Fri, 19 Aug 2016 21:47:29 +0000 (05:47 +0800)]
uv: dont handle SIGSEGV, SIGFPE
https://github.com/warmcat/libwebsockets/issues/609
Patrick Gansterer [Sun, 14 Aug 2016 12:04:56 +0000 (14:04 +0200)]
Remove context from lws_context_per_thread for non-libuv builds
The variable is never read when built without libuv.
Patrick Gansterer [Sun, 14 Aug 2016 10:51:15 +0000 (12:51 +0200)]
Add error handling for SSL_new() of clients
Do not access wsi->ssl if SSL_new() failed and log the error.
Patrick Gansterer [Sun, 14 Aug 2016 10:38:03 +0000 (12:38 +0200)]
Fix signature of lws_create_vhost() in README.coding.md
Patrick Gansterer [Sat, 13 Aug 2016 09:17:53 +0000 (11:17 +0200)]
Allow serving files without known mimetype
RFC2616 only says that any HTTP/1.1 message containing an entity-body
SHOULD include a Content-Type header field defining the media type of
that body.
RFC2119 defines SHOULD as: This word mean that there may exist valid
reasons in particular circumstances to ignore a particular item, but
the full implications must be understood and carefully weighed before
choosing a different course.
AG: this isn't an oversight, it's paranoia about sending out /etc/passwd
or /etc/shadow accidentally.
I agree it should be allowed if people really want to override it. But
the default should remain like it is I think.
I adapted the patch to allow the extra mimetype "*": "" to be declared on
a mount, as a wildcard match that serves the file without a Content-Type.
Andy Green [Wed, 10 Aug 2016 21:36:08 +0000 (05:36 +0800)]
server max protocol element 64
https://github.com/warmcat/libwebsockets/issues/601
Andy Green [Mon, 8 Aug 2016 13:54:30 +0000 (21:54 +0800)]
client-add-PUSH-http-body-capability
Support HTTP body sending on client connections.
Add demo to test-client.
Run the normal test server, then
$ libwebsockets-test-client http://localhost:7681/formtest -o
It will post the string "hello" to the POST test using application/x-www-form-urlencoded
https://github.com/warmcat/libwebsockets/issues/598
Also ensures any of the Client HTTP transient modes will call back LWS_CALLBACK_CLOSED_CLIENT_HTTP
if they close
https://github.com/warmcat/libwebsockets/issues/600
Andy Green [Sun, 7 Aug 2016 00:33:08 +0000 (08:33 +0800)]
client fixups after esp8266
Andy Green [Sat, 23 Jul 2016 06:18:25 +0000 (14:18 +0800)]
esp8266 initial support
Andy Green [Wed, 3 Aug 2016 19:18:36 +0000 (03:18 +0800)]
base64 decode fix lengths
Andy Green [Tue, 19 Jul 2016 01:38:48 +0000 (09:38 +0800)]
ah detach: don't print held time if actually detached
Confusingly an ah held time was printed for a previously detached ah.
Clear down the time field when detaching the ah so this can't happen any more.
Andy Green [Fri, 15 Jul 2016 05:41:38 +0000 (13:41 +0800)]
ws ping pong on idle connections
This adds a new member to the context creation info struct "ws_ping_pong_interval".
If nonzero, it sets the number of seconds that established ws connections are
allowed to be idle before a PING is forced to be sent. If zero (the default) then
tracking of idle connection is disabled for backwards compatibility.
Timeouts cover both the period between decision to send the ping and it being
sent (because it needs the socket to become writeable), and the period between
the ping being sent and the PONG coming back.
INFO debug logs are issues when the timeout stuff is operating.
You can test the server side by running the test server hacked to set ws_ping_pong_interval
and debug log mask of 15. Both the mirror protocol and the server-status protocol are
idle if nothing is happening and will trigger the PING / PONG testing. (You can also
test using lwsws and /etc/lwsws/conf with "ws-pingpong-secs": "20" in the global section)
For client, run the test client with -n -P 20 for 20s interval. -n stops the test client
writing using the mirror protocol, so it will be idle and trigger the PING / PONGs.
The timeout interval may be up to +10s late, as lws checks for affected connections every
10s.
Andy Green [Wed, 13 Jul 2016 03:35:16 +0000 (11:35 +0800)]
Plugin server-status cleanup
Andy Green [Wed, 13 Jul 2016 00:45:22 +0000 (08:45 +0800)]
plugins-dim-webpage-when-connection-lost
Andy Green [Sat, 25 Jun 2016 22:29:20 +0000 (06:29 +0800)]
plugin table dirlisting
Andy Green [Wed, 13 Jul 2016 00:14:57 +0000 (08:14 +0800)]
doxygen use sections
Mike Messina [Mon, 11 Jul 2016 13:17:48 +0000 (21:17 +0800)]
Recheck slot when an FD gets closed and make formatting consistent
mmessina [Mon, 11 Jul 2016 13:17:21 +0000 (21:17 +0800)]
Use one event for all sockets to avoid 64 limit and fix the single dispatch issue
Andy Green [Mon, 11 Jul 2016 01:44:17 +0000 (09:44 +0800)]
SNI-vhost-matching-fallback-to-wildcard
Andy Green [Sun, 10 Jul 2016 23:48:53 +0000 (07:48 +0800)]
ah change default header len to 4096 and pool size 4
Users are starting to appear with clients sending more than the default max header
content buffer of 1024... with the advent of the ah pool the old reasons for keeping this
modest no longer apply, so up it to 4096 and reduce the default pool size to 4 from 16 to
keep the overall memory usage the same.
Andy Green [Sun, 10 Jul 2016 23:28:23 +0000 (07:28 +0800)]
EXTPOLL export LWS_POLLXXX and document .events must be .revents
Andy Green [Thu, 7 Jul 2016 00:14:26 +0000 (08:14 +0800)]
client ssl add flag to control server cert hostname check
Andy Green [Wed, 6 Jul 2016 09:42:19 +0000 (17:42 +0800)]
defer ESTABLISHED until after mode is WS_SERVING
Andy Green [Wed, 6 Jul 2016 02:48:41 +0000 (10:48 +0800)]
lejp-conf: substitute _lws_ddir_ with install dir
Andy Green [Tue, 5 Jul 2016 23:37:34 +0000 (07:37 +0800)]
lejp_conf: pmo as alias for cgi-env
Since cgi-env only applies for cgi types, we can
re-use it for generic per-mount options (pmo)
Andy Green [Tue, 5 Jul 2016 10:40:40 +0000 (18:40 +0800)]
win32-clean-and-use-do-while
Andy Green [Sun, 3 Jul 2016 01:20:11 +0000 (09:20 +0800)]
lws_get_mimetype
Andy Green [Sat, 2 Jul 2016 01:23:15 +0000 (09:23 +0800)]
peer closes during partial buffered just hang up
https://github.com/warmcat/libwebsockets/issues/573
Andy Green [Fri, 1 Jul 2016 00:54:39 +0000 (08:54 +0800)]
CLIENT_CONNECTION_ERROR add strings
This clears up a couple of issues with client connect.
- if CLIENT_CONNECTION_ERROR is coming, which of the many
ways the rejection may have happened is documented in the
in argument. It's still possible if it just got hung up on
in will be NULL, but now it has MANY more canned strings
describing the issue available at the callback
"getaddrinfo (ipv6) failed"
"unknown address family"
"getaddrinfo (ipv4) failed"
"set socket opts failed"
"insert wsi failed"
"lws_ssl_client_connect1 failed"
"lws_ssl_client_connect2 failed"
"Peer hung up"
"read failed"
"HS: URI missing"
"HS: Redirect code but no Location"
"HS: URI did not parse"
"HS: Redirect failed"
"HS: Server did not return 200"
"HS: OOM"
"HS: disallowed by client filter"
"HS: disallowed at ESTABLISHED"
"HS: ACCEPT missing"
"HS: ws upgrade response not 101"
"HS: UPGRADE missing"
"HS: Upgrade to something other than websocket"
"HS: CONNECTION missing"
"HS: UPGRADE malformed"
"HS: PROTOCOL malformed"
"HS: Cannot match protocol"
"HS: EXT: list too big"
"HS: EXT: failed setting defaults"
"HS: EXT: failed parsing defaults"
"HS: EXT: failed parsing options"
"HS: EXT: Rejects server options"
"HS: EXT: unknown ext"
"HS: Accept hash wrong"
"HS: Rejected by filter cb"
"HS: OOM"
"HS: SO_SNDBUF failed"
"HS: Rejected at CLIENT_ESTABLISHED"
- until now the user code did not get the new wsi that was created
in the client connection action until it returned. However the
client connection action may provoke callbacks like
CLIENT_CONNECTION_ERROR before then, if multiple client connections
are initiated it makes it unknown to user code which one the callback
applies to. The wsi is provided in the callback but it has not yet
returned from the client connect api to give that wsi to the user code.
To solve that there is a new member added to client connect info struct,
pwsi, which lets you pass a pointer to a struct wsi * in the user code
that will get filled in with the new wsi. That happens before any
callbacks could be provoked, and it is updated to NULL if the connect
action fails before returning from the client connect api.
Andy Green [Thu, 30 Jun 2016 02:21:23 +0000 (10:21 +0800)]
remove polarssl mbedtls support
Just remove it from cmake
Andy Green [Thu, 30 Jun 2016 02:11:59 +0000 (10:11 +0800)]
reduce log spew and document test-server variants
Robin Rowe [Wed, 29 Jun 2016 18:02:08 +0000 (02:02 +0800)]
fix __x86_64__ check
https://github.com/warmcat/libwebsockets/issues/574
Andy Green [Tue, 28 Jun 2016 22:04:25 +0000 (06:04 +0800)]
test-libev add server-status protocol
Andy Green [Tue, 28 Jun 2016 11:50:40 +0000 (19:50 +0800)]
client confirm server hostname in cert
Openssl v1.0.2 and above have support for checking the hostname
the client side connected to against the hostname on the cert the
server presented.
This enables that feature if the necessary API is available in the
openssl version, meaning the connection will fail at ssl negotiation if the
cert isn't for the requested server
It's very easy to test, add a fake entry to /etc/hosts for the server IP with
a different name, using that will fail at ssl but using the correct dns name
matching the certificate will work.
Andy Piper [Tue, 28 Jun 2016 11:01:20 +0000 (19:01 +0800)]
windows detect client connection error
Andy Green [Sun, 26 Jun 2016 21:53:38 +0000 (05:53 +0800)]
dlfcn h only if plugins
Signed-off-by: None <andy@warmcat.com>
Andy Green [Mon, 20 Jun 2016 09:05:31 +0000 (17:05 +0800)]
documentation convert to doxygen
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Wed, 15 Jun 2016 04:24:38 +0000 (12:24 +0800)]
protocol-lws-messageboard
This is a simple messageboard built on top of lwsgs
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Sat, 18 Jun 2016 01:00:04 +0000 (09:00 +0800)]
lws_bind_protocol
When using http/1.1+ keepalive and mounts, the relationship between
a connection and a protocol becomes dynamic. The same connection might
visit different bits of the url space served by different mounts using
different protocols.
This patch ensures protocols can cleanly manage their per-connection
allocations by using the following callbacks when the protocol changes
LWS_CALLBACK_HTTP_BIND_PROTOCOL
LWS_CALLBACK_HTTP_DROP_PROTOCOL
For example if the pss wants to malloc stuff at runtime, it should do it
in LWS_CALLBACK_HTTP_BIND_PROTOCOL or later, and clean it up in
...DROP_PROTOCOL.
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Thu, 19 May 2016 07:28:31 +0000 (15:28 +0800)]
protocol generic sessions
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Fri, 17 Jun 2016 02:05:23 +0000 (10:05 +0800)]
lws_json_escape
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Fri, 17 Jun 2016 22:36:37 +0000 (06:36 +0800)]
lejp conf report human readable errors
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Fri, 17 Jun 2016 01:41:22 +0000 (09:41 +0800)]
lws_now_secs
Signed-off-by: Andy Green <andy@warmcat.com>
OndraCo [Thu, 16 Jun 2016 05:02:31 +0000 (13:02 +0800)]
Modified the 64 connections "backup" so that there is no wait if any event
is already ready.
OndraCo [Thu, 16 Jun 2016 05:02:20 +0000 (13:02 +0800)]
Added a back-up in case of more than 64 connections.
Andy Green [Wed, 15 Jun 2016 02:46:58 +0000 (10:46 +0800)]
lws_callback_vhost_protocols
This gives protocols a way to talk to each other via per-vhost callbacks,
one per protocol (including the sender).
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Sun, 12 Jun 2016 01:56:39 +0000 (09:56 +0800)]
lws_process_html_args
Signed-off-by: Andy Green <andy@warmcat.com>
Andy Green [Wed, 8 Jun 2016 02:07:02 +0000 (10:07 +0800)]
introduce urlencode decode and sql escape public apis
This adds
- simple lws_urlencode()
- simple lws_urldecode()
- simple lws_sql_purify
Those expect the data to all be there and process it up until
the first '\0'.
There is also a larger opaque apis for handling POST_BODY urldecode. To
enable these, you need to give cmake -DLWS_WITH_STATEFUL_URLDECODE=1 (or
arrange any larger feature that relies on it sets that in CMakeLists.txt)
- stateful urldecode with parameter array
These have create / process / destroy semantics on a struct that maintains
decode state.
Stateful urldecode is capable of dealing with large POST data in multiple
POST_BODY callbacks cleanly, eg, file transfer by POST.
Stateful urldecode with parameter array wraps the above with a canned
callback that stores the urldecoded data and indexes them in a pointer
array matching an array of parameter names.
You may also pass it an optional callback when creating it, that will recieve
uploaded file content.
The test html is updated to support both urlencoded and multipart forms,
with some javascript to do clientside validation of an arbitrary 100KB
file size limit (there is no file size limit in the apis).
Signed-off-by: Andy Green <andy@warmcat.com>
Gadkari Mugdha [Mon, 13 Jun 2016 09:23:44 +0000 (17:23 +0800)]
fix for https connection code