This rips out the connection hashtable implementation along with
MAX_CLIENTS and replaces it with a dynamically allocated fds array
and lookup table along the same lines as the new extpoll implementation
from Edwin van den Oetelaar.
It detects the max number of file descriptors possible at context init
time and allocates accordingly; this can be externally controlled by
ulimit and the server run as a specific user to facilitate targeting
specific ulimit rules at it.
Many operations that translated between socket descriptors and struct
websocket or pollfd objects have had iteration removed by this patch
and under load will be a lot faster.
Signed-off-by: Andy Green <andy.green@linaro.org>
From an idea by Edwin van den Oetelaar <oetelaar.automatisering@gmail.com>
When testing libwebsockets with ab, Edwin found an unexpected bump in
the distribution of latencies, some connections were held back almost
the whole test duration.
http://ml.libwebsockets.org/pipermail/libwebsockets/2013-January/000006.html
Studying the problem revealed that when there are mass pending connections
amongst many active connections, we do not service the listen socket often
enough to clear the backlog, some seem to get stale violating FIFO ordering.
This patch introduces listen socket service "piggybacking", where every n
normal socket service actions we also check the listen socket and deal with
pending connections there.
Normally, it checks the listen socket gratuitously every 10 normal socket
services. However, if it finds something waiting, it forces a check on the
next normal socket service too by keeping stats on how often something was
waiting. If the probability of something waiting each time becomes high,
it will allow up to two waiting connections to be serviced for each normal
socket service.
In that way it has low burden in the normal case, but rapidly adapts by
detecting mass connection loads as found in ab.
Signed-off-by: Andy Green <andy.green@linaro.org>
Default remains at SOMAXCONN, you can force it at configure time
along these lines
./configure CFLAGS="-DLWS_SOMAXCONN=16384"
Signed-off-by: Andy Green <andy.green@linaro.org>
Previously we sat and looped to dump a file over http protocol.
Actually that's a source of blocking to the other sockets being serviced.
This patch breaks up the file service into a roundtrip around the poll()
loop for each 512-byte packet. It doesn't make much difference if the
server is idle, but if it's busy it makes sure everyone else is getting
service while the file is sent.
It doesn't try to optimize multiple users of the file or to keep the
descriptor open, the point of this patch is to establish the breaking up
of the file send action into the poll loop.
On the user side, there are two differences:
- context is now needed in the first argument to libwebsockets_serve_http_file()
that's not too bad since we provide context in the callback.
- file send is now asynchronous to the user code, you get a new callback coming
in protocol 0 when it's done, LWS_CALLBACK_HTTP_FILE_COMPLETION
libwebsockets-test-server is updated accordingly.
Signed-off-by: Andy Green <andy.green@linaro.org>
"4b0e01f Retry SSL_connect when SSL_get_error requests it. " from David Galeano
noticed the problem that client connect may receive SSL_ERROR_WANT_* from
SSL_connect, which is basically WOULDBLOCK. That patch tried to deal with it
by blocking in a while(1) until the condition went away.
That's problematic because of it blocks service of anything else (including
the host application sockets in the external socket poll sharing case) for
up to 5s controlled by conditions at one client.
After fiddling with and researching this, the actual problem with the code is
we are not getting the SSL layer error correctly, it is not contained in the
code returned from the Connect api directly.
I was unable to get a renegotiation forced on my modern SSL libs, it complained
about protocol error are reopened the connection instead. So I think the stuff
found in the docs and the web about the SSL_ERROR_WANT_ is probably not something
we will see in reality (if we check the right error code...)
Signed-off-by: Andy Green <andy.green@linaro.org>
This patch allows control of the main compiletime constants in libwebsockets
from the configure commandline.
README is updated with documentation on what's available, how to set them
and the defaults.
The constants are logged with "info" severity (not visible by default) at
context create time.
The zlib constant previously exposed like this is moved to private-libwebsockets.h
so it can be printed along with the rest.
Signed-off-by: Andy Green <andy.green@linaro.org>
- multiple debug context calls lwsl_ err, warn, debug, parser, ext, client
- api added to set which contexts output to stderr using a bitfield log_level
- --disable-debug on configure removes all code that is not err or warn severity
- err and warn contexts always output to stderr unless disabled by log_level
- err and warn enabled by default in log_level
Signed-off-by: Andy Green <andy@warmcat.com>
This patch gets deflate-stream working with x-google-mux.
It adds a clean veto system where are extension can veto the proposal
of any extension when opening a new connection. x-google-mux uses that
in its callback to defeat any use of deflate-stream on mux children.
However deflate stream is allowed on the parent connection and works
transparently now alongside x-google-mux.
Signed-off-by: Andy Green <andy@warmcat.com>
This is initial x-google-mux support. It's disabled by default
since it's very pre-alpha.
1) To enable it, reconfigure with --enable-x-google-mux
2) It conflicts with deflate-stream, use the -u switch on
the test client to disable deflate-stream
3) It deviates from the google standard by sending full
headers in the addchannel subcommand rather than just
changed ones from original connect
4) Quota is not implemented yet
5) Close of subchannel is not really implemented yet
6) Google opcode 0xf is changed to 0x7 to account for
v7 protocol changes to opcode layout
However despite those caveats, in fact it can run the
test client reliably over one socket (both dumb-increment
and lws-mirror-protocol), you can open a browser on the
same test server too and see the circles, etc.
Signed-off-by: Andy Green <andy@warmcat.com>
Mozilla implementationcan issue window of up to 15,
need to match it
Reported-by: Patrick McManus <pmcmanus@mozilla.com>
Signed-off-by: Andy Green <andy@warmcat.com>
Ago noticed that some Windows clients experience small packets
from the server being aggregated and set after a long delay
(200-300ms).
He found that TCP_NODELAY on the socket solved this, I tested it
and it didn't have any noticable bad effect, so I implemented it
for all sockets, client and server.
Thans Ago for debugging this and notifying the cause.
Reported-by: Ago Allikmaa <maxorator@gmail.com>
Signed-off-by: Andy Green <andy@warmcat.com>
This adds win32 build compatability to libwebsockets.
The patch is from Peter Hinz, Andy Green has cleaned it up a bit and
possibly broken win32 compatability since I can't test it, so there
may be followup patches. It compiles fine under Linux after this
patch anyway.
Much of the patch is changing a reserved keyword for Visual C compiler
"this" to "context", but there is no real C99 support in the MSFT
compiler even though it is 2011 so C99 style array declarations
have been mangled back into "ancient C" style.
Some windows-isms are also added like closesocket() but these are
quite localized. Win32 random is just using C library random() call
at the moment vs Linux /dev/urandom. canonical hostname detection is
broken in win32 at the moment.
Signed-off-by: Peter Hinz <cerebusrc@gmail.com>
Signed-off-by: Andy Green <andy@warmcat.com>
Doing a client connect was atomic until now, blocking
all the other service while it waited for proxy and / or
server response.
This patch uses the new timeout system and breaks the
client connect sequence into three states handled by
the normal poll() processing. It means that there are
now no blocking network delays and it's all handled
by the main state machine.
Signed-off-by: Andy Green <andy@warmcat.com>
This adds a concept of timeouts for operations enforced by
connection closure if the timeout is reached.
Once a second all sockets are checked for timing out, every time
there is a service call it checks to see if a second has passed since
the last check and checks if so.
You can also call libwebsocket_service_fd() with a NULL fd to give
the timeouts a chance to be detected; if it's less than a second since
the last check it returns immediately.
Signed-off-by: Andy Green <andy@warmcat.com>
