Adds a new api lws_vhost_destroy(struct lws_vhost *) which allows dynamic removal of vhosts.
The external api calls two parts of internal helpers that get reused for context destroy.
The second part is called deferred by 5s... this is to ensure that event library objects
composed into structs owned by the vhost all have a chance to complete their close
asynchronously. That should happen immediately, but it requires us to return to the
event loop first.
The vhost being removed is deleted from the context vhost list by the first part, and does
not block further removals or creation during the delay for the deferred freeing of the
vhost memory.
Part 1:
- if the vhost owned a listen socket needed by other vhosts listening on same iface + port, the listen
socket is first handed off to another vhost so it stays alive
- all wsi still open on the vhost are forcibly closed (including any listen socket still attached)
- inform all active protocols on the vhost they should destroy themselves
- remove vhost from context vhost list (can no longer be found by incoming connections)
- add to a "being destroyed" context list and schedule the second part to be called in 5s
Part 2:
- remove us from the being destroyed list
- free all allocations owned by the vhost
- zero down the vhost and free the vhost itself
In libwebsockets-test-server, you can send it a SIGUSR1 to have it toggle the creation and destruction of
a second vhost on port + 1.
Sometimes "Assertion failed: !uv__is_closing(handle)" happens because
handle is being closed multiple times. To fix this, "uv_is_closing"
is added before calling "uv_close".
Signed-off-by: Petar Paradzik <petar.paradzik@sartura.hr>
1) This makes lwsws run a parent process with the original permissions.
But this process is only able to respond to SIGHUP, it doesn't do anything
else.
2) You can send this parent process a SIGHUP now to cause it to
- close listening sockets in existing lwsws processes
- mark those processes as to exit when the number of active connections
on the falls to zero
- spawn a fresh child process from scratch, using latest configuration
file content, latest plugins, etc. It can now reopen listening sockets
if it chooses to, or open different listen ports or whatever.
Notes:
1) lws_context_destroy() has been split into two pieces... the reason for
the split is the first part closes the per-vhost protocols, but since
they may have created libuv objects in the per-vhost protocol storage,
these cannot be freed until after the loop has been run.
That's the purpose of the second part of the context destruction,
lws_context_destroy2().
For compatibility, if you are not using libuv, the first part calls the
second part. However if you are using libuv, you must now call the
second part from your own main.c after the first part.
This should allow adding vhosts "late", ie, after the server is up and
running with its initial vhost(s). The necessary housekeeping is folded
into lws_create_vhost() itself so it should be transparent.
Notice though that at the point the server starts to do service after it
starts initially, if it was requested that the UID / GID change, that
is performed at that point and is not reversible.
So vhosts added "late" find themselves running under the unprivileged
UID / GID from the very start, whereas vhosts added "early" initially
run under the UID / GID the process started with. If protocols the
vhost uses want to, eg, open privileged files at init and then use
them unprivileged, that will fail if the vhost is added late because
the initial privs are already gone.
AG: also deal with lws_protocol_init() on late vhost init (does the
callbacks for per vh protocol creation), add comments
Via Dosvald
lws_service_tsi() which has been around a while actually just
calls through to lws_plat_service_tsi(), meaning there is no
need to expose both apis.
Rename the internal lws_plat_service_tsi() to _lws_plat_service_tsi()
and replace the api export with a #define to lws_service_tsi for
compatibility's sake.
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.
If OOT lws plugins will be packaged as separate projects,
they're going to want to install their plugins somewhere
that makes sense for the package instead of one big lws
plugin dir.
This patch changes info to have a const char ** to a NULL
terminated array of directories it should search for
plugins. lwsws knows about this and you can add to the
dir array using config fragments like
{
"global": {
"plugin-dir": "/usr/local/share/coherent-timeline/plugins"
}
}
if the config fragment in /etc/lwsws/conf.d/ is also managed by the
package with the plugin, it can very cleanly add and remove itself
from lwsws based on package install status.
Signed-off-by: Andy Green <andy@warmcat.com>
This trades off a couple of wsi pointers for vastly increased speed
for the callback when writeable "all protocol" variants when there
are many kinds of wsi active.
Signed-off-by: Andy Green <andy@warmcat.com>
After discussion here
https://libwebsockets.org/pipermail/libwebsockets/2016-April/002268.html
scandir usage in
- lwsws conf.c
- lws plugin support
and
- lws plugin apis for dl
are converted to us libuv apis so they can work cross-platform easily.
lws itself remains not requiring libuv, although it's an option.
- LWS_WITH_LWSWS
- LWS_WITH_PLUGINS
now force LWS_WITH_LIBUV if selected... both of these are new features
only in master atm and both are off by default in CMake.
There's a complication libuv can be too old to offer the necessary apis,
this is the case in Travis Trusty instance. In that case, UV_VERSION_MAJOR ==0,
then the unix-only plugin implementation is used instead.
Signed-off-by: Andy Green <andy@warmcat.com>
This adds support for dynamically loaded plugins at runtime, which
can expose their own protocols or extensions transparently.
With these changes lwsws defaults to OFF in cmake, and if enabled it
automatically enables plugins and libuv support.
Signed-off-by: Andy Green <andy@warmcat.com>
This patch splits out some lws_context members into a new lws_vhost struct.
- ssl state and options per vhost
- SSL_CTX for serving and client per vhost
- protocols[] per vhost
- extensions[] per vhost
lws_context maintains a linked list of lws_vhosts.
The same lws_context_creation_info struct is used to regulate both the
context creation and to create vhosts: for backward compatibility if you
didn't provide the new LWS_SERVER_OPTION_EXPLICIT_VHOSTS option, then
a default vhost is created at context creation time using the same info
data as the context itself.
If you will have multiple vhosts though, you should give the
LWS_SERVER_OPTION_EXPLICIT_VHOSTS option at context creation time,
create the context first and then the vhosts afterwards using
lws_create_vhost(contest, &info);
Although there is a lot of housekeeping to implement this change, there
is almost no additional overhead if you don't use multiple vhosts and
very little api impact (no changes to test apps).
Signed-off-by: Andy Green <andy@warmcat.com>
https://github.com/warmcat/libwebsockets/issues/468
Adds lws_check_opt() to regularize multibit flag checking.
There's a new context creation flag LWS_SERVER_OPTION_DO_SSL_GLOBAL_INIT,
this is included automatically if you give any other SSL-related option flag.
If you give no SSL-related option flag, nor this one directly, then even
though SSL support may be compiled in, it is never initialized nor used for the
whole lifetime of the lws context.
Conversely in order to prepare the context to use SSL, even though, eg, you
are not listening on SSL but will use SSL client connections later, you can
give this flag explicitly to make sure SSL is initialized.
Signed-off-by: Andy Green <andy@warmcat.com>
This gets the libuv stuff plumbed in and working.
Currently it's only workable for some service thread, and there
is an isolated valgrind problem left
==28425== 128 bytes in 1 blocks are definitely lost in loss record 3 of 3
==28425== at 0x4C28C50: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==28425== by 0x4C2AB1E: realloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==28425== by 0x58BBB27: maybe_resize (core.c:748)
==28425== by 0x58BBB27: uv__io_start (core.c:787)
==28425== by 0x58C1B80: uv__signal_loop_once_init (signal.c:225)
==28425== by 0x58C1B80: uv_signal_init (signal.c:260)
==28425== by 0x58BF7A6: uv_loop_init (loop.c:66)
==28425== by 0x4157F5: lws_uv_initloop (libuv.c:89)
==28425== by 0x405536: main (test-server-libuv.c:284)
libuv wants to sign off on all libuv 'handles' that will close, and
callback to do the close confirmation asynchronously. The wsi close function
is adapted when libuv is in use to work with libuv accordingly and exit the uv
loop the number of remaining wsi is zero.
Signed-off-by: Andy Green <andy.green@linaro.org>
