Joakim Soderberg [Fri, 22 Feb 2013 01:27:57 +0000 (09:27 +0800)]
Bugfix compiling for cross compiling.
! does not mean NOT, use NOT instead :)
Larry Hayes [Fri, 22 Feb 2013 01:16:20 +0000 (09:16 +0800)]
keepalive swap interval and probes
The interval and probes members for setting keepalive options
per-connection on Linux were swapped.
Signed-off-by: Larry Hayes <lhayesg@gmail.com>
Edwin van den Oetelaar [Wed, 20 Feb 2013 12:56:59 +0000 (20:56 +0800)]
provide socketfd at in param for LWS_CALLBACK_FILTER_NETWORK_CONNECTION
Signed-off-by: Edwin van den Oetelaar <oetelaar.automatisering@gmail.com>
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Wed, 20 Feb 2013 11:11:31 +0000 (19:11 +0800)]
trac 18 deal with service connection timing out
While looking at http://libwebsockets.org/trac/ticket/18
noticed the flow for timeout in service_fd will do bad things
if the fd we came to service has timed out. It gets freed and
then "serviced'.
Reported-by: Joakim Soderberg <joakim.soderberg@gmail.com>
Signed-off-by: Andy Green <andy.green@linaro.org>
Edwin van den Oetelaar [Tue, 19 Feb 2013 18:08:51 +0000 (02:08 +0800)]
cmake fix for older cmake
AG this wasn't needed on cmake 2.8.10.2 but did not make any
problem either
Signed-off-by: Edwin van den Oetelaar <oetelaar.automatisering@gmail.com>
Andy Green [Tue, 19 Feb 2013 11:19:51 +0000 (19:19 +0800)]
update missed extpoll calls to use correct args for ssl
http://libwebsockets.org/trac/ticket/15#comment:16
Reported-by: Joakim Soderberg <joakim.soderberg@gmail.com>
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 19 Feb 2013 02:26:39 +0000 (10:26 +0800)]
cmake docs add note about prefix
Signed-off-by: Andy Green <andy.green@linaro.org>
Joakim Söderberg [Tue, 19 Feb 2013 02:14:54 +0000 (10:14 +0800)]
Create the "doc" dir before generating docs.
The documentation wouldn't be generated if the path didn't exist.
Joakim Söderberg [Tue, 19 Feb 2013 02:14:37 +0000 (10:14 +0800)]
Added "make install" support to the CMake project.
- Does everything as the autoconf script does (I think).
- Generates SSL certs for the test-server
- Installs all test apps.
- Defaults to /usr/local/
- "DESTDIR=/bla make install" works
- Append "libwebsockets" to start of the test-app names like the autoconf stuff does.
- Only tested on OSX so far...
Andy Green [Tue, 19 Feb 2013 02:01:48 +0000 (10:01 +0800)]
trac 17 update set_log_level api notice
http://libwebsockets.org/trac/ticket/17
Reported-by: dbrnz
Signed-off-by: Andy Green <andy.green@linaro.org>
Edwin van den Oetelaar [Mon, 18 Feb 2013 12:20:05 +0000 (20:20 +0800)]
move cms cmake addition to cmake module paths earlier
Signed-off-by: Edwin van den Oetelaar <oetelaar.automatisering@gmail.com>
Andy Green [Mon, 18 Feb 2013 10:34:10 +0000 (18:34 +0800)]
add FindGit.cmake
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 08:54:26 +0000 (16:54 +0800)]
document ensure_user_space going private
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 08:30:10 +0000 (16:30 +0800)]
remove lws_ensure_user_space from public api change return
The function has a logical problem when the size of the requested
allocation is 0, it will return NULL which is overloaded as
failure.
Actually the whole function is evil as an api, this patch moves
it out of the public API space and fixes it to return 0 for
success or 1 for fail. Private code does not need to to return
wsi->user_space and public code should only get that from the
callback as discussed on trac recently.
Thanks to Edwin for debugging the problem.
Reported-by: Edwin van den Oetelaar <oetelaar.automatisering@gmail.com>
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 04:08:15 +0000 (12:08 +0800)]
update arm build stats
Clean out the historical numbers, update info
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 04:02:18 +0000 (12:02 +0800)]
trim proxy ads plus hostname stg
128 should be fine for hostname and proxy ads
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 03:32:49 +0000 (11:32 +0800)]
restrict http send buffer to 4096
send() is able to complete but not send everything asked...
we don't deal with that yet. On an ARM9 board, leaf.jpg
failed because the send() only accepted 7240 of 8192... not
a failure but it implies reporting partial send to the
caller. Work around by keeping the send buffer at 4096.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 02:43:18 +0000 (10:43 +0800)]
fix no extensions build
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 02:38:45 +0000 (10:38 +0800)]
migrate client hs and c_port into ah
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 02:34:25 +0000 (10:34 +0800)]
minor type optimizations
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 02:29:22 +0000 (10:29 +0800)]
remove current_alloc_len
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 02:22:42 +0000 (10:22 +0800)]
remove header name buffer
The header name buffer and its max length handling has actually
been unused since the minilex parser was introduced. We hold
parsing state in the lex-type parts and don't need to store or
worry about max length, since the parser will let us know as
soon as it can't be a match for the valid header names.
This strips it out reducing the per-connection allocation for
x86_64 with default configure from 224 to 160.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 18 Feb 2013 01:48:31 +0000 (09:48 +0800)]
handle http union transition properly
There's no proper transition to http union state until now.
It only becomes apparant there's a problem when you try to
return -1 from the HTTP callback, during the close action
it will try to close() a nonsense, nonzero fd pointer in
the uninitialized u.http union member.
This patch takes a copy of the allocated headers struct from
the u.hdr union state, transitions to u.http clearing down u
and then calls the HTTP callback with URI args pointing to
the still-in-scope ah allocation. After the call, the copy ah
is freed.
That makes sure we are in the correct union state while still
giving the HTTP callback access to the URI without having to
copy it around.
Reported-by: Edwin can den Oetelaar <oetelaar.automatisering@gmail.com>
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 17 Feb 2013 01:14:08 +0000 (09:14 +0800)]
test server kill skt with minus 1
The client spams the server with mirror protocol content
and then closes the connection. However the server wants to
send that content back to all the mirror protocol
connections, including the one that closed.
The result is during the send back phase we usually see an
error trying to send to the close client connection. Because
we don't return -1, we don't close it on server side and the
error repeats a lot before finally closing. This is a side-
effect of a recent patch to drain pending input before
really closing a dead connection.
This patch just changes it to return -1 on the failed send
attempt assertively closing the connection.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sat, 16 Feb 2013 02:17:52 +0000 (10:17 +0800)]
handle any POLLIN before error
See if this can impact http://libwebsockets.org/trac/ticket/16
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Fri, 15 Feb 2013 14:48:58 +0000 (22:48 +0800)]
close if we tried to close politely just close next time
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Fri, 15 Feb 2013 14:36:30 +0000 (22:36 +0800)]
api change deliver socket fd to in param of extpoll callbacks
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Fri, 15 Feb 2013 14:32:53 +0000 (22:32 +0800)]
problems in lws_handle_POLLOUT_event should just close
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Fri, 15 Feb 2013 14:31:55 +0000 (22:31 +0800)]
robustness protect and document ensure_user_space
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Fri, 15 Feb 2013 02:25:58 +0000 (10:25 +0800)]
README.coding add note about closing connections
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 15:06:37 +0000 (23:06 +0800)]
add by hand http send example
This adds a demonstration of how to use the http nonblocking
send action to the test server.
If you ask for /leaf.jpg from the test server, it will send
"by hand" a 2.4MB jpeg in HTTP, including the headers. See
the test server sources to see how it is done.
Although it's a file, and a jpeg image, actually the exact
same scheme will work for any data or mime type.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 15:18:10 +0000 (23:18 +0800)]
create user alloc for http if requested
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 14:23:54 +0000 (22:23 +0800)]
introduce LWS_CALLBACK_HTTP_WRITEABLE
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 09:11:22 +0000 (17:11 +0800)]
just get hostname into canonical_hostname
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 06:15:58 +0000 (14:15 +0800)]
fragge able to send chunks larger than rx buffer
This now exercises the code for dealing with rx that spills over
the rx buffer. Single fraggle uses the default rx buffer size
of 4096, it also now exercises code around that.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 03:25:44 +0000 (11:25 +0800)]
unstaged server changes
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 03:23:49 +0000 (11:23 +0800)]
update 1.21 changelog
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Thu, 14 Feb 2013 02:18:31 +0000 (10:18 +0800)]
check for default protocol rx buf limit
This fixes
http://libwebsockets.org/trac/ticket/13
When using the default rx protocol buffer, the check is
performed against 0 not the default length. That's the
case both in client and server code...
There's no problem if you actually give a max frame size
in the protocol definition.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Wed, 13 Feb 2013 03:54:08 +0000 (11:54 +0800)]
update changelog
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Wed, 13 Feb 2013 03:44:20 +0000 (11:44 +0800)]
fix missing cr from closing log
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Wed, 13 Feb 2013 01:34:53 +0000 (09:34 +0800)]
update echo to use externsion getting api
Signed-off-by: Andy Green <andy.green@linaro.org>
Joakim Soderberg [Wed, 13 Feb 2013 01:29:26 +0000 (09:29 +0800)]
Fixed DLL compilation on Windows for CMake.
- Define LWS_DLL and LWS_INTERNAL when websockets_shared is compiled.
- The websocket_shared target compiles to websocket.lib / websocket.dll
(websocket.lib contains the exported functions for websocket.dll, and is
the file that is linked to when a program wants to use the dll)
- The websocket target compiles to websocket_static.lib on windows.
- Replaced any "extern" with "LWS_EXTERN" on libwebsockets.h for proper
DLL function exports.
- Created a LIB_LIST with all the libwebsocket dependencies, instead of
multiple calls to target_link_libraries, only one call is made for both
the static and shared library version. This makes it easy to add other
variants if wanted in the future.
- Added ZLIB as a dependency for the libs, so that the build order will be
correct at all times.
- Added a dependency for the websockets lib to the test apps, so it is
built before them.
- Fixed the test-server-extpoll app to include the emulated_poll, and link
to winsock on Windows.
- Removed the global export of libwebsocket_internal_extensions, and added
a function libwebsocket_get_internal_extensions() that returns it
instead. Using the global would not work with the DLL export on Windows.
Joakim Soderberg [Wed, 13 Feb 2013 01:29:17 +0000 (09:29 +0800)]
Fixed soname and build shared lib for CMake.
- Build a shared version of the library (used to be only static).
- Set the so version name properly since the ABI has changed for version 1.2
Joakim Soderberg [Wed, 13 Feb 2013 01:29:09 +0000 (09:29 +0800)]
Added so-version information to the lib.
Andy Green [Wed, 13 Feb 2013 01:08:43 +0000 (09:08 +0800)]
disable fstack usage
It's only workable on gcc 4.6+... and in fact it's only
going to be interesting typically to the maintainer, so it's
commented out in Makefile.am.
Currently the worst remaining ones for x86_64 are
libwebsockets.c:2250:6:_lws_log 496 static
libwebsockets.c:2203:13:lwsl_emit_stderr 352 static
client.c:700:1:libwebsockets_generate_client_handshake 320 static
server.c:42:1:libwebsockets_decode_ssl_error 288 static
server-handshake.c:31:1:handshake_0405 272 static
client.c:368:1:lws_client_interpret_server_handshake 272 static
libwebsockets.c:447:1:libwebsockets_get_peer_addresses 256 static
daemonize.c:93:1:lws_daemonize 240 static
output.c:536:5:libwebsockets_serve_http_file 224 static
sha-1.c:131:1:sha1_step 192 static
sha-1.c:316:1:SHA1 160 static
libwebsockets.c:1604:1:libwebsocket_create_context 160 static
The top two there are error assembly buffers, they could be made
static but then they won't be usable from threaded user code.
Reported-by: Anders Brander <anders@brander.dk>
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 14:44:43 +0000 (22:44 +0800)]
echo test app needs different lockfile
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 07:48:00 +0000 (15:48 +0800)]
update rpm specfile
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 04:56:05 +0000 (12:56 +0800)]
introduce attack script
Seems like it would be a good idea to try to mess with the
server at least before someone else does it for us
Just run the script
$ test-server/attack.sh
it will spawn a test server and fire things at it. If you
see the end result
---- survived
then you should be OK.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 05:10:19 +0000 (13:10 +0800)]
security disallow repeated GET
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 04:52:39 +0000 (12:52 +0800)]
security harden http parser a bit
Drop the connection during parsing for a few more cases that can't be legit.
Take care about trying to free rxflow_buffer only if we reached a connmode
where it exists
Change behaviour on setting unknown HTTP method to kill connection
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 04:53:36 +0000 (12:53 +0800)]
fix another escaape runon
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 02:50:49 +0000 (10:50 +0800)]
add cyassl keepalive valgrind minimal mem to changelog
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 02:36:39 +0000 (10:36 +0800)]
bump version to 1.2 and soname to 3
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 02:19:08 +0000 (10:19 +0800)]
api make close api private only
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 02:15:25 +0000 (10:15 +0800)]
api remove hangup_on_client
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 02:10:32 +0000 (10:10 +0800)]
update test echo for iface info member namechange
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 02:07:22 +0000 (10:07 +0800)]
eliminate snprintf
The two cases where I introduced snprintf are either already
safe for buffer overflow or can be made so with one extra
statement, allowing sprintf.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Tue, 12 Feb 2013 02:13:02 +0000 (10:13 +0800)]
fix info struct api docs for iface vs interface
Signed-off-by: Andy Green <andy.green@linaro.org>
Joakim Soderberg [Mon, 11 Feb 2013 16:52:23 +0000 (17:52 +0100)]
Fixed compilation on Windows.
Andy Green [Mon, 11 Feb 2013 14:40:39 +0000 (22:40 +0800)]
update memory performance info
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 13:43:41 +0000 (21:43 +0800)]
fix ssl reject path for freeing header allocation
If the SSL connection failed before the headers came, we were not
dealing with deallocating the header malloc. This takes care of it.
Using CyaSSL, we are then valgrind-clean for ssl client and server.
With OpenSSL, there is 88 bytes lost at init that never changes or
gets recovered. AFAIK there's nothing to do about that.
OpenSSL also blows these during operation
==1059== Conditional jump or move depends on uninitialised value(s)
==1059== at 0x4A0B131: bcmp (mc_replace_strmem.c:935)
==1059== by 0x3014CDDBA8: ??? (in /usr/lib64/libcrypto.so.1.0.1c)
==1059== by 0x3015430852: tls1_enc (in /usr/lib64/libssl.so.1.0.1c)
==1059== by 0x3015428CEC: ssl3_read_bytes (in /usr/lib64/libssl.so.1.0.1c)
==1059== by 0x30154264C5: ??? (in /usr/lib64/libssl.so.1.0.1c)
==1059== by 0x4C3C596: lws_server_socket_service (server.c:153)
==1059== by 0x4C32C1E: libwebsocket_service_fd (libwebsockets.c:927)
==1059== by 0x4C33270: libwebsocket_service (libwebsockets.c:1225)
==1059== by 0x401C84: main (in /usr/bin/libwebsockets-test-server)
However googling around
https://bugs.launchpad.net/ubuntu/+source/openssl/+bug/60021
http://www.openssl.org/support/faq.html#PROG13
(also the next FAQ down)
it seems OpenSSL have a relaxed attitude to this and it's expected.
It's interesting CyaSSL works fine but doesn't have that problem...
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 12:10:56 +0000 (20:10 +0800)]
fix busted debug format in ssl mode
CyaSSL blows (symptomless?) uninitialized memory accesses in
valgrind when using SSL_get_version()... don't need to do it...
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 12:03:59 +0000 (20:03 +0800)]
fix string escape runon
Oh well it was a big patch to just be cosmetic ^^
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 11:36:15 +0000 (19:36 +0800)]
fix without server
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 09:13:32 +0000 (17:13 +0800)]
style cleanup
This brings the library sources into compliance with checkpatch
style except for three or four exceptions like WIN32 related stuff
and one long string constant I don't want to break into multiple
sprintf calls.
There should be no functional or compilability change from all
this (hopefully).
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 06:50:45 +0000 (14:50 +0800)]
valgrind openssl destroy as far as possible
OpenSSL doesn't seem to have a way to close out three allocations
it creates during SSL library init.
Even after doing the magic incantations suggested on the openssl list, we're
left with these. Well, 88 bytes from ssl init is probably not critical,
but it's annoying
==15206== HEAP SUMMARY:
==15206== in use at exit: 88 bytes in 3 blocks
==15206== total heap usage: 13,566 allocs, 13,563 frees, 5,933,134 bytes allocated
==15206==
==15206== 24 bytes in 1 blocks are still reachable in loss record 1 of 3
==15206== at 0x4A06409: malloc (vg_replace_malloc.c:270)
==15206== by 0x3014C612B2: CRYPTO_malloc (in /usr/lib64/libcrypto.so.1.0.1c)
==15206== by 0x3015441B38: ??? (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x3015443A78: SSL_COMP_get_compression_methods (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x301544932B: SSL_library_init (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x4C340D4: libwebsocket_create_context (libwebsockets.c:1796)
==15206== by 0x401C08: main (in /usr/bin/libwebsockets-test-server)
==15206==
==15206== 32 bytes in 1 blocks are still reachable in loss record 2 of 3
==15206== at 0x4A06409: malloc (vg_replace_malloc.c:270)
==15206== by 0x3014C612B2: CRYPTO_malloc (in /usr/lib64/libcrypto.so.1.0.1c)
==15206== by 0x3014CC91BE: sk_new (in /usr/lib64/libcrypto.so.1.0.1c)
==15206== by 0x3015441AF9: ??? (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x3015443A78: SSL_COMP_get_compression_methods (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x301544932B: SSL_library_init (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x4C340D4: libwebsocket_create_context (libwebsockets.c:1796)
==15206== by 0x401C08: main (in /usr/bin/libwebsockets-test-server)
==15206==
==15206== 32 bytes in 1 blocks are still reachable in loss record 3 of 3
==15206== at 0x4A06409: malloc (vg_replace_malloc.c:270)
==15206== by 0x3014C612B2: CRYPTO_malloc (in /usr/lib64/libcrypto.so.1.0.1c)
==15206== by 0x3014CC91DC: sk_new (in /usr/lib64/libcrypto.so.1.0.1c)
==15206== by 0x3015441AF9: ??? (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x3015443A78: SSL_COMP_get_compression_methods (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x301544932B: SSL_library_init (in /usr/lib64/libssl.so.1.0.1c)
==15206== by 0x4C340D4: libwebsocket_create_context (libwebsockets.c:1796)
==15206== by 0x401C08: main (in /usr/bin/libwebsockets-test-server)
==15206==
==15206== LEAK SUMMARY:
==15206== definitely lost: 0 bytes in 0 blocks
==15206== indirectly lost: 0 bytes in 0 blocks
==15206== possibly lost: 0 bytes in 0 blocks
==15206== still reachable: 88 bytes in 3 blocks
==15206== suppressed: 0 bytes in 0 blocks
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 06:32:02 +0000 (14:32 +0800)]
valgrind client close in a controlled way on SIGINT
With this, both the test server and client are valgrind-clean
tested with a chrome session also active to the server
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 06:32:48 +0000 (14:32 +0800)]
valgrind eliminate uninitialized warning on close
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 06:12:32 +0000 (14:12 +0800)]
valgrind also deallocate rx buf on close when client
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 06:08:50 +0000 (14:08 +0800)]
valgrind dont close things directly in the callback
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 06:05:02 +0000 (14:05 +0800)]
valgrind client go through context destroy on connection error
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 05:04:45 +0000 (13:04 +0800)]
client convert to new headers scheme eliminating mallocs
This removes all the direct wsi members specific to clients,
most of them are moved to being fake headers in the next 3-layer
header scheme, c_port moves to being a member of the u.hdr
unionized struct.
It gets rid of a lot of fiddly mallocs and frees(), despite it
adds a small internal API to create the fake headers, actually
the patch deletes more than it adds...
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 04:37:13 +0000 (12:37 +0800)]
valgrind free rx_user_buffer if entered CONNMODE_WS_SERVING
This seems to be enough to get a clean valgrind run for the
test server with 1 x chrome and 1 x libwebsockets-test-client
session being run for 10s
lwsts[19767]: libwebsockets-test-server exited cleanly
==19767==
==19767== HEAP SUMMARY:
==19767== in use at exit: 0 bytes in 0 blocks
==19767== total heap usage: 41,071 allocs, 41,071 frees, 27,464,834 bytes allocated
==19767==
==19767== All heap blocks were freed -- no leaks are possible
==19767==
==19767== For counts of detected and suppressed errors, rerun with: -v
==19767== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 2 from 2)
test client is another story...
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 04:05:54 +0000 (12:05 +0800)]
valgrind introduce protocol init and destroy user callbacks
To get a clean bill of health from valgrind, we have to have a way to
inform the user code that we're going down and it should free everything
it is holding that was malloc'd.
This patch introduces LWS_CALLBACK_PROTOCOL_DESTROY which each protocol
gets when the context is being destroyed and no more activity will come
after that call. They can get rid of everything there.
To match it, LWS_CALLBACK_PROTOCOL_INIT is introduced which would allow
one-time init per protocol too.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 03:43:05 +0000 (11:43 +0800)]
valgrind context destroy close all conns properly
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 03:27:44 +0000 (11:27 +0800)]
valgrind drop header allocation down http path
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 10:02:31 +0000 (18:02 +0800)]
replace per header mallocs with single malloc 3 level struct
This big patch replaces the malloc / realloc per header
approach used until now with a single three-level struct
that gets malloc'd during the header union phase and freed
in one go when we transition to a different union phase.
It's more expensive in that we malloc a bit more than 4Kbytes,
but it's a lot cheaper in terms of malloc, frees, heap fragmentation,
no reallocs, nothing to configure. It also moves from arrays of
pointers (8 bytes on x86_64) to unsigned short offsets into the
data array, (2 bytes on all platforms).
The 3-level thing is all in one struct
- array indexed by the header enum, pointing to first "fragment" index
(ie, header type to fragment lookup, or 0 for none)
- array of fragments indexes, enough for 2 x the number of known headers
(fragment array... note that fragments can point to a "next"
fragment if the same header is spread across multiple entries)
- linear char array where the known header payload gets written
(fragments point into null-terminated strings stored in here,
only the known header content is stored)
http headers can legally be split over multiple headers of the same
name which should be concatenated. This scheme does not linearly
conatenate them but uses a linked list in the fragment structs to
link them. There are apis to get the total length and copy out a
linear, concatenated version to a buffer.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 03:04:56 +0000 (11:04 +0800)]
improve static allocation notice
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 03:04:01 +0000 (11:04 +0800)]
valgrind free context allocations
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 02:03:00 +0000 (10:03 +0800)]
remove extension cruft from struct lws
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 01:37:23 +0000 (09:37 +0800)]
use part of service buffer to make response not malloc
Done with an offset because the encoded key is stored at the
start of service_buffer at this time
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 01:32:53 +0000 (09:32 +0800)]
remove minimum frame size for deflate
The idea here seems wrong, if we have a mixture of frames of varying
sizes above and below the limit, we segfault in deflate after skipping
it once.
If the protocol doesn't want compression because many frames are
small, it should veto the extension in the user callback. If only
a few frames are tiny, the overhead for compressing it all is tiny.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Mon, 11 Feb 2013 01:31:43 +0000 (09:31 +0800)]
stop O2 override
Now we are building with -O0 -g and debug enabled by default.
--disable-debug in configure will get you a -04 without -g
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 13:21:24 +0000 (21:21 +0800)]
dont close in user callback wrapper let ancestor do it
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 12:58:04 +0000 (20:58 +0800)]
fix error path in file transfer
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 08:25:20 +0000 (16:25 +0800)]
throw out lws_websocket_related cruft
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 08:00:47 +0000 (16:00 +0800)]
optimize wsi using bitfields and enums to chars
Also max protocols to 5 (overridable by configure) and max extensions
from 10 to 3 by default (also overridable by configure).
wsi is down to 608 on x86_64 with this.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 07:34:59 +0000 (15:34 +0800)]
use context service buffer instead of stack for clent_connect
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 07:19:39 +0000 (15:19 +0800)]
use context service buffer instead of stack for clent_connect_2
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 07:10:10 +0000 (15:10 +0800)]
use context service buffer instead of stack for create_context
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 14:22:01 +0000 (22:22 +0800)]
fix non ssl changes missed from context api change
http://libwebsockets.org/trac/ticket/11#comment:5
Reported-by: amn
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 03:03:32 +0000 (11:03 +0800)]
use context service buffer instead of stack for lws_client_socket_service
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 02:43:04 +0000 (10:43 +0800)]
use context service buffer instead of stack for server_socket_service
Reduces lws_server_socket_service from 4208 to 80 stack allocation
on x86_64
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 02:23:24 +0000 (10:23 +0800)]
add static stack analysis
You can get a worst-first list of stack allocators like this
$ cat ./lib/.libs/*.su | sort -k2g | tac
On x86_64, currently the ones above 100 bytes are
server.c:126:5:lws_server_socket_service 4208 static
client.c:42:5:lws_client_socket_service 1584 static
libwebsockets.c:1539:1:libwebsocket_create_context 1136 static
libwebsockets.c:783:1:libwebsocket_service_fd 656 static
client-handshake.c:3:22:__libwebsocket_client_connect_2 592 static
libwebsockets.c:2149:6:_lws_log 496 static
server-handshake.c:33:1:handshake_0405 464 static
libwebsockets.c:2102:13:lwsl_emit_stderr 352 static
client.c:351:1:lws_client_interpret_server_handshake 240 static
daemonize.c:93:1:lws_daemonize 224 static
libwebsockets.c:434:1:libwebsockets_get_peer_addresses 208 static
client.c:694:1:libwebsockets_generate_client_handshake 208 static
output.c:534:5:libwebsockets_serve_http_file 192 static
output.c:51:6:lwsl_hexdump 176 static
sha-1.c:316:1:SHA1 160 static
libwebsockets.c:157:1:libwebsocket_close_and_free_session 144 static
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 02:46:45 +0000 (10:46 +0800)]
dont try figure out listen_service_fd position if unset
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 01:39:47 +0000 (09:39 +0800)]
dont try set per socket keepalive timing on bsds
As per http://libwebsockets.org/trac/ticket/10
BSD doesn't support setting keepalive info per-socket
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sun, 10 Feb 2013 01:06:38 +0000 (09:06 +0800)]
fix broken listen socket piggybacking
As reported here
http://libwebsockets.org/trac/ticket/11
the code for connection storm handling had rotted, fds[0] is no longer
always related to the listen socket when serving.
This patch updates it to determine the listen socket correctly to stop
infinite recursion here.
Reported-by: amn
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sat, 9 Feb 2013 04:25:31 +0000 (12:25 +0800)]
introduce keepalive option and make common socket options function
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sat, 9 Feb 2013 06:20:17 +0000 (14:20 +0800)]
remove receiving timeout for client
Now we enforce nonblocking everywhere, hopefully properly,
this rx timeout stuff shouldn't be needed.
Signed-off-by: Andy Green <andy.green@linaro.org>
Andy Green [Sat, 9 Feb 2013 06:10:04 +0000 (14:10 +0800)]
correct test client to close synchronously with last send
Noticed previously the test client lifetime is decoupled from
the actual send activity. With SSL if the connection starts but
there is a period of SSL-layer "blocking" (actually fail-and-retry)
the mirror lifetime could be exhausted before the connection really
completed, making it stall after it was then closed.
This corrects that so connection lifetime management is done in the
send callback.
Signed-off-by: Andy Green <andy.green@linaro.org>