On Thu, Sep 15, 2011 at 10:15 PM, Gaspard Bucher
<gaspard@teti.ch> wrote:
I have this strange case where up-values in a coroutine are garbage collected but the coroutine is not (and when it runs, the userdata has 0x0 pointers).
Here is my dependency graph (~~~ = weak value) :
scheduler ---> wrap ---> t ~~~~~> thread ---> coroutine --- (up-value) --> foobar ---> super (userdata)
From my understanding, when "foobar" or "thread" goes out of scope, the whole thing should be garbage collected. What is happening is that only the
userdata inside foobar is collected but foobar is still a table. This table does not have any special mode.
getmetatable(foobar)
=> nil
Any clues on this would be greatly appreciated..
Cheers,
This is still failing... When I add a debug objects (prints a string on gc):
-------------- ~D (debug object garbage collected)
<lk.Debug: 0x0> (I can do a print on the debug object)
As we can see above, the debug object has been garbage collected but I can still print it which means the object is accessible through the coroutine and should not be collected.
How is this possible ?
--- We should provide a socket to inform when registration is over (callback).
-- This socket could be the default zmq.REQ socket used to by lk.Service ?
function mdns.Registration(service_type, name, port, func)
func = func or dummy
local self = constr(service_type, name, port)
self.thread = lk.Thread(function()
local d = lk.Debug("D")
while true do
sched:waitRead(self:fd())
print(d)
func(self:getService())
end
end)
return self
end