Submitted: Fabrice Fontaine ( ffontaine ) - 2012-03-05 01:55:54 PST
element was not freed if membuffer_append_str(&url_str, "http://") does
not return 0. Moreover addNew was not used.
Submitted: Fabrice Fontaine ( ffontaine ) - 2012-03-05 01:31:16 PST
Memsetting to 0 some of the structures: finfo in process_request,
job in readFromSSDPSocket, request in http_OpenHttpGetEx, job in
genaNotifyThread, job in genaNotifyAllExt, job in genaNotifyAll,
job in genaInitNotifyExt, job in genaInitNotify, LocalAddr in
getlocalhostname.
Submitted: Fabrice Fontaine ( ffontaine ) - 2012-03-05 01:15:34 PST
An out-of-bands access is raised because size of
RespInst->AcceptLanguageHeader is 200 and TmpBuf size is 180.
Submitted: Fabrice Fontaine ( ffontaine ) - 2012-02-29 02:09:43 PST
newAttrNode is not freed if newAttr->n.nodeValue = strdup(value); returns
NULL or if ixmlElement_setAttributeNodeNS(element, newAttr, NULL) does
not return IXML_SUCCESS.
Submitted: Fabrice Fontaine ( ffontaine ) - 2012-02-28 04:27:54 PST
ixmlFreeDOMString(ActNodeName); is called twice if
ixmlParseBufferEx(ActNodeName, RespNode); does not return IXML_SUCCESS.
Submitted: Fabrice Fontaine ( ffontaine ) - 2012-02-28 04:01:02 PST
There is a memory leak in ixmlDocument_createElementEx:
newElement->tagName is not freed if
newElement->n.nodeName = strdup(tagName); returns NULL.
Submitted: Marcelo Roberto jimenez ( mroberto ) - 2012-02-26 16:50:23 PST
src/genlib/net/http/httpparser.c: In function ‘parser_parse_requestline’:
src/genlib/net/http/httpparser.c:1319:28: warning: ‘index’ may be used uninitialized in this function
(cherry picked from commit 9125d82010)
Submitted: Yoichi NAKAYAMA ( yoichi ) - 2012-02-21 07:06:35 PST
In genaSubscribe() (defined in upnp/src/gena/gena_ctrlpt.c),
ActualSID and EventURL will not be freed if ScheduleGenaAutoRenew
returns UPNP_E_SUCCESS.
This fixes the an issue introduced by the previous fix.
Submitted: Yoichi NAKAYAMA ( yoichi ) - 2012-02-21 07:06:35 PST
In genaSubscribe() (defined in upnp/src/gena/gena_ctrlpt.c),
ActualSID and EventURL will not be freed if ScheduleGenaAutoRenew
returns UPNP_E_SUCCESS.
This fixes the original issue.
Submitted by Berend Dekens ( Berend Dekens ) - 2012-02-08 06:24:31 PST
In httpparser.c on line 1385 it says that HTTP 1.0 replies are blocked
because the UPnP verfication tool requires this.
I looked in the specs and as far as I can find, one should only be
carefull to send chunked communication to hosts supporting HTTP 1.1.
There is no requirement to support only HTTP 1.1.
The XBMC media server uses the Platinum UPnP library which replies
using HTTP/1.0 messages. As it is now, libupnp returns an error while
trying to parse the response while the response itself is completely
valid.
Is there a requirement in the UPnP 1.0 spec that I missed or is this
restriction self-imposed? And can it be lifted?
-------------------------------------------------------------------
Comment by Fabrice Fontaine:
Hide
Hi,
You're right, this modification should be removed. This version checking
was wrongly added in parser_parse_responseline function.
...
Attached is a patch that resolved an issue I found with a server that
gave its device description URI without a trailing slash (e.g.
`http://127.0.0.1:5555`).
Small bug fix on IPv6 Control Point: now CP will also send M-SEARCH on
site-scope address (FF05::C) instead of only sending M-SEARCH on
link-scope (FF02::C).
SF Bug tracker, ID: 3469344
Submitted: dimmman ( dimmman ) - 2012-01-04 01:44:29 PST
Details: Looking at the code (v1.6.14, upnptools.c) for UpnpResolveURL
and UpnpResolveURL2 it shows that the ExitFunction: always returns
UPNP_E_SUCCESS.
I'm farily sure it's a simple mistake that should have been "return ret;"
in both cases.
Br,
Jonny
Adding two new functions (UpnpSendAdvertisementLowPower and
UpnpUnRegisterRootDeviceLowPower) which can be used to specify values
for the three SSDP headers defined by UPnP Low Power. Those headers are
Powerstate, SleepPeriod and RegistrationState.
Putting all access to ssdpReqSock4 and ssdpReqSock6 under
INCLUDE_CLIENT_APIS compilation flag to be able to compile when
client part of library is disable.
Add a new UpnpRegisterRootDevice4 which allow user to specify a
description URL to be returned for legacy CPs (for example, CPs
searching for a v1 when the device is v2). Most of those CPs does not
work if they found a v2 in the XML description, so this new function is
only used to solve interoperability issues.
Legacy applications like linux-igd and igd2-for-linux are using those
API to create a thread pool for managing their GENA events.
Leave it to be reworked in 1.8.x.
Major bug fix in miniserver.c for IPv6, bug was introduced when
changing implementation of get_port in November 20th 2010 ("gena:fix
several compiler warnings" commit).
The files FreeList.h, LinkedList.h, ThreadPool.h and TimerThread.h
from the threautil library were being installed in the include
directory of the library, incorrectly exposing internal data structure
of the library.
Add 'requiredThreads' field to the ThreadPool structure, to avoid
a race condition when waiting for a new thread to be created. The
race condition occurs when a thread is destroyed while the master
thread is waiting for a new thread to be created.
Thanks to Chuck Thomason for pointing the problem.
Summary: Race condition can hang miniserver thread - ID: 3158591
Details:
Hello,
I have found a race condition in the thread pool handling of
libupnp-1.6.6 that periodically results in the miniserver thread
getting blocked infinitely.
In my setup, I have the miniserver thread pool configured with 1
job per thread, 2 threads minimum, and 50 threads maximum.
Just before the lockup occurs, the miniserver thread pool contains
2 threads: one worker thread hanging around from a previous HTTP
request job (let's call that thread "old_worker") and the
miniserver thread itself.
A new HTTP request comes in. Accordingly, the miniserver enters
schedule_request_job() and then ThreadPoolAdd(). In
ThreadPoolAdd(), the job gets added to the medium-priority queue,
and AddWorker() is called. In AddWorker(), jobs = 1 and threads =
1, so CreateWorker gets called.
When we enter CreateWorker(), tp->totalThreads is 2, so
currentThreads is 3. The function creates a new thread and then
blocks on tp->start_and_shutdown. The miniserver thread expects
the newly created thread to increment tp->totalThreads and then
signal the condition variable to wake up the miniserver thread and
let it proceed.
The newly created thread starts in the WorkerThread() function. It
increments tp->totalThreads to 3, does a broadcast on the
start_and_shutdown condition, and starts running its job. However,
before the miniserver thread wakes up, "old_worker" times out. It
sees that there are no jobs in any queue and that the total number
of threads (3) is more than the minimum (2). As a result, it
reduces tp->totalThreads to 2 and dies.
Now the miniserver thread finally wakes up. It checks
tp->totalThreads and sees that its value is 2, so it blocks on
tp->start_and_shutdown again. It has now "missed" seeing
tp->totalThreads get incremented to 3 and will never be unblocked
again.
When this issue does occur for a server device, the miniserver
port remains open, but becomes unresponsive since the miniserver
thread is stuck. SSDP alive messages keep getting sent out, as
they are handled by a separate thread. Reproducing the issue is
difficult due to the timing coincidence involved, but in my
environment I am presently seeing it at least once a day. I
figured out the sequence described above through addition of my
own debug logs.
The relevant code involved in this bug has not changed
substantially in libupnp-1.6.10, though I am planning to test
against 1.6.10 as well in the near future.
Do you have any input for an elegant fix for this issue?
Thanks,
Chuck Thomason
