Adding a configure flag to disable GENA notification reordering as even
with an imillisleep(1), this mechanism consumes too much CPU on embedded
devices when there is a burst of notifications.
When a device with embedded devices (like IGD) when created and one of
the embedded devices did not have any service, there was a Segmentation
Fault (see SF Tracker [ 2688125 ]).
When a lot of notifications were generated by a device in a short
period of time then 100% of the CPU was used to reorder those
notifications by pushing back the thread in the job queue. This
mechanism has been modified so now thread sleep 1 ms before being
pushed back into the job queue.
Removing DEFAULT_SCHED_PARAM parameter and use
sched_get_priority_min(DEFAULT_POLICY) instead.
Devices must respond to M-SEARCH requests for any supported version and the
response should specify the same version as was contained in the search target.
Previously, the device did not answer if the M-SEARCH request did not
contain the same version number than the version number of the device.
This patch adds the WEB_SERVER_CONTENT_LANGUAGE parameter so the user can specify
the language used by the device during Description and Presentation steps of UPnP
through the HTTP CONTENT-LANGUAGE header.
By default, the WEB_SERVER_CONTENT_LANGUAGE is an empty string so no
CONTENT-LANGUAGE is added.
This patch allows a user to customize the stack size of the threads used by
pupnp through the new THREAD_STACK_SIZE variable. This is especially useful
on embedded systems with limited memory where the user can set THREAD_STACK_SIZE
to ITHREAD_STACK_MIN.
However, as this modification can have side effects, I set 0 as the default
value, so threads will continue to use the default stack size of the system
(which varies greatly as stated in
https://computing.llnl.gov/tutorials/pthreads/).
Hi,
I've made some modification to the libupnp v1.6.5
I've add a timeout for each TCP connect.
It is very useful when an UPnP device stop working and do not accept
connection for an UPnP action.
Modifications are only located in
upnp/src/genlib/net/http/httpreadwrite.c
For every TCP connection, I set the socket to non-blocking, perform
connect,
check result and wait during a timeout if necessary, then reset the
socket to blocking.
Please see this patch in attached file.
I hope it helps.
Regards,
David Hoeung
Consultant Extia
Orange Labs R&D
----
I discovered a reliable denial-of-service issue on the last stable
release of libupnp (1.6.6) remotely triggerable by any
unauthenticated user. The issue is related with a bad parsing of
malformed XML.
This is a follow up from issue 6 in tracker id 3056713: calling UpnpSetMaxContentLength() by passing '0' disables the content length checking. This is useful for developing some prototype applications that deal with a lot of XML/SOAP data, and for debugging.
The corresponding c file change is already in the pupnp tree. Copy/pasting the relevant block of code here for clarity:
In upnp/src/genlib/net/http/httpreadwrite.c:
if (g_maxContentLength > 0 && parser->content_length > (unsigned int)g_maxContentLength) {
*http_error_code = HTTP_REQ_ENTITY_TOO_LARGE;
line = __LINE__;
ret = UPNP_E_OUTOF_BOUNDS;
goto ExitFunction;
}
This block of code checks only does the bounds check if g_maxContentLength > 0, and it's only place g_maxContentLength is checked.
Attached is a patch against the latest sources.
(cherry picked from commit 7f1e164a5a)
src/ssdp/ssdp_ctrlpt.c: In function ‘SearchByTarget’:
src/ssdp/ssdp_ctrlpt.c:634: warning: format ‘%s’ expects type ‘char *’, but argument 6 has type ‘int’
Sending messages over UDP is broken in some Apple OSes
such as OS X and iOS. This might be broken in other OSes to but didn't
verify.
The fix is to modify the socket lenght argument of sendto to use the correct
sockaddr lenght dependng on whether the socket is IPV4 or IPV6.
Also added some error checks and debugging related to the issue
crash. This happens when the file being downloaded exceeds the device
memory - entirely possible when transferring video files.
The programmatic cause is that the logic implemented in the function
http_ReadHttpGet (which UpnpReadHttpGet calls) reads the entire file
into memory. The fix modifies the existing logic to discard data after
it's been read; there's no reason to keep it around since the caller
of UpnpReadHttpGet already has a copy of it.
This issue exists in 1.6.6 as well as the latest sources.
Patch submitted by Chandra (inactiveneurons).
use connect() are broken. More specifically, connect() in these methods
is returning with an EINVAL. The programatic cause is that the address_len
argument passed to connect() is different in IPV4 vs IPV6 (as described in:
http://www.opengroup.org/onlinepubs/009695399/functions/connect.html).
The current code always uses the IPV6 size. The fix modifies each use of
connect() to use the correct size based on the address family being used.
Patch submitted by Chandra (inactiveneurons).
