Files
ServerCore/SubscriptionManager.cpp
T
Matt Arant 3383595296 Fix unsynchronized data race in SubscriptionManager::processCommand()
processCommand() read/mutated the `subscriptions` map and individual
Subscription objects' `subscribers` vectors (via subscribe()/unsubscribe()/
event()/the "publish" branch's own insert()) without taking `lock` at all,
even though add() and removeSessionSubscriptions() already do take it for
the exact same state. removeSessionSubscriptions() runs from ~TCPSession(),
on whatever thread happens to be tearing down a *different* connection, so
one thread erasing a session from a Subscription's subscribers vector could
run concurrently with another thread iterating (event()) or appending to
(subscribe()) that same vector - a plain unsynchronized data race.

Reproduced live while testing an unrelated ItemDataBase fix: rapid connect/
subscribe/disconnect cycles against the "createItem" subscription produced
a real segfault, confirmed via gdb - Socket::write() -> resetSocket()
dereferencing a corrupted EPoll reference, consistent with iterating a
vector that another thread was concurrently erasing from.

Wrapped the whole method body in a std::lock_guard on the same `lock`
mutex SubscriptionManager already has. Verified no reentrancy risk: nothing
reachable from processCommand() (including subscription handlers' own
process()/onSubscribe() implementations) calls back into processCommand(),
add(), or removeSessionSubscriptions() on the same thread, and `lock` is a
plain non-recursive std::mutex, so this can't deadlock.

Verified fix: the exact repro sequence that segfaulted the server twice
now runs clean, server stays up.

Not touched: Subscription::sendToAll() has no live callers (dead code);
dynamically `new`'d *OWNER subscriptions are still never delete()'d when
unpublished (a pre-existing leak, not a crash) - both out of scope here.
2026-09-07 20:18:14 +00:00

224 lines
8.3 KiB
C++

#include "SubscriptionManager.h"
#include "Log.h"
#include "Subscription.h"
#include "SubscriptionHandlerFactory.h"
#include "TCPServer.h"
#include <algorithm>
#include <mutex>
namespace core
{
int SubscriptionManager::add(Subscription &subscription)
{
lock.lock();
subscriptions.insert(std::make_pair(subscription.id, &subscription));
lock.unlock();
return 1;
}
bool SubscriptionManager::onClearSubscription(coreutils::MString temp, coreutils::MString key)
{
temp = "";
key = "";
return true;
}
int SubscriptionManager::removeSessionSubscriptions(TCPSession &session)
{
int countSubscribed = 0;
int countPublished = 0;
int count = 0;
lock.lock();
coreutils::MString temp = "";
for (auto [key, subscription] : subscriptions)
{
if (temp != "")
{
subscriptions.erase(temp);
temp = "";
}
countSubscribed += subscription->unsubscribe(session);
if (subscription->subscriptionOwners == coreutils::MString(session.alias))
{
if (subscription->selectionOfOwnerForSubscription == "*DIE")
{
onClearSubscription(temp, key);
}
else if (subscription->will == "*NEXT")
{
if (subscription->subscribers.size() < 1)
{
coreutils::Log(coreutils::LOG_DEBUG_1) << "There is not enough people to move to the next";
onClearSubscription(temp, key);
}
else
{
for (auto subscribed = subscription->subscribers.begin(); subscribed < subscription->subscribers.end(); subscribed++)
{
if ((*subscribed)->alias == session.alias)
{
subscription->subscribers.erase(subscribed++);
}
subscription->subscriptionOwner(session.alias, "*NEXT");
}
}
}
else
{
coreutils::Log(coreutils::LOG_DEBUG_1) << "Subscription doesn't exist";
}
++countPublished;
}
}
if (temp != "")
{
subscriptions.erase(temp);
temp = "";
}
coreutils::Log(coreutils::LOG_DEBUG_2) << "Removed session from " << countSubscribed << " subscription(s).";
coreutils::Log(coreutils::LOG_DEBUG_2) << "Cancelled " << countPublished << " channel(s) for session.";
lock.unlock();
return countSubscribed;
}
int SubscriptionManager::processCommand(coreutils::ZString &request, TCPSession &session)
{
// Every branch below reads or mutates either the `subscriptions` map
// or some Subscription's own `subscribers` vector - both of which
// removeSessionSubscriptions() (called from ~TCPSession(), on
// whatever thread happens to be tearing down a *different*
// connection) also reads/mutates, under `lock`. This method used to
// take no lock at all, so a session disconnecting on one thread
// could run subscribers.erase() concurrently with subscribe()/
// event() iterating or push_back()-ing the same vector on another -
// an unsynchronized data race. Confirmed live: rapid connect/
// subscribe/disconnect cycles against ItemDataBase's "createItem"
// subscription produced a real segfault (gdb: Socket::write() ->
// resetSocket() dereferencing a corrupted EPoll reference) while
// broadcasting to *SUBSCRIBERS. `lock` is a plain, non-recursive
// std::mutex, and nothing reachable from this function re-enters
// processCommand()/add()/removeSessionSubscriptions() on the same
// thread, so holding it for the whole call is safe.
std::lock_guard<std::mutex> guard(lock);
if (request[0] == "publish")
{
SubscriptionHandler *handler = NULL;
if (request.getList().size() > 3)
{
handler = factory->getSubscriptionHandler(request[4].str());
}
if (request[2] == "*OWNER")
{
newSubscription = new Subscription(request[1].str(), session.alias, request[2].str(), request[3].str(), handler, request[5].str(), request[6].str());
subscriptions.insert(std::make_pair(request[1].str(), newSubscription));
return 1;
}
return 1;
}
else if (request[0] == "catalog")
{
session.out << ":catalog:";
for (auto const &[key, subscription] : subscriptions)
{
session.out << subscription->id << ";";
}
session.out << std::endl;
return 1;
}
else if (request[0] == "invite")
{
std::stringstream out;
coreutils::Log(coreutils::LOG_DEBUG_1) << request[2];
std::string invitee = request[2].str();
// Was getSessionByAlias(&invitee) - getSessionByAlias(void*)
// compares by POINTER IDENTITY against each session's own
// alias, but &invitee is the address of a local variable that
// was just created in this function call - it can never equal
// any real session's alias, so this lookup was guaranteed to
// return NULL every time, and the result was dereferenced
// unconditionally below. Any client sending a bare "invite:"
// crashed the entire process, not just its own connection -
// same class of bug as the two GuildManager crashes fixed
// this session. This generic ServerCore layer has no
// name-based session lookup available (that only exists as a
// Zephora-specific addition on top of TCPServer), so a real
// fix for "find a session by the player name they typed"
// isn't possible at this layer without a larger change; this
// stops the crash without pretending to fix that gap.
TCPSession *tempSession = session.server->getSessionByAlias(&invitee);
if (tempSession == NULL)
{
coreutils::Log(coreutils::LOG_DEBUG_1) << "invite: could not find session for " << invitee;
return 0;
}
std::stringstream temp;
temp << "invite:" << request[1] << ":" << invitee;
tempSession->write(temp.str());
return 1;
}
// Was subscriptions[request[1].str()] - map::operator[] default-
// constructs (here: NULL) and INSERTS an entry for any id that
// isn't already registered, and subscription->id below was then
// dereferenced completely unconditionally. Any subscribe:/
// unsubscribe:/event: naming an id nobody ever created (confirmed
// live: "subscribe:invite", sent by the client on every scene
// load, referencing an "invite" subscription that plain doesn't
// exist anywhere - real invites are Publish()'d under a
// per-group-guid id, not a shared "invite" channel) null-derefs
// and takes down the ENTIRE process, every connection included,
// not just the one that sent it. find() avoids polluting the map
// with the bogus NULL entry too.
auto it = subscriptions.find(request[1].str());
if (it == subscriptions.end())
{
coreutils::Log(coreutils::LOG_DEBUG_1) << "Unknown subscription id: " << request[1].str();
return 0;
}
auto subscription = it->second;
if (request[1].equals(subscription->id))
{
if (request[0] == "unpublish")
{
subscriptions.erase(request[1].str());
}
else if (request[0] == "subscribe")
{
subscription->subscribe(session);
return 1;
}
else if (request[0] == "unsubscribe")
{
subscription->unsubscribe(session);
return 1;
}
else if (request[0] == "event")
{
std::stringstream out;
subscription->process(request, out, session);
if (subscription->mode == "*ANYONE")
{
subscription->event(out);
return 1;
}
else if (subscription->mode == "*SUBSCRIBERS")
{
if (subscription->ifSubscriber(session))
{
subscription->event(out);
return 1;
}
}
}
return 0;
}
return 0;
}
}