Memory leak under Windows

[jswirl]

This is after applying this PR: #48
I'm not really familiar with Windows, so I could be doing something wrong, but it appears there may be leaks some place else.

The Windows UMDH utility seems to point to some suspicious parts in sctp_init_ifns_for_vrf():
https://github.com/sctplab/usrsctp/blob/master/usrsctplib/netinet/sctp_bsd_addr.c#L356-L362 and
https://github.com/sctplab/usrsctp/blob/master/usrsctplib/netinet/sctp_bsd_addr.c#L404-L409.

Here's the UMDH log from the test program I used:
stackdiff.txt

[tuexen]

OK. Can you retest with the current code? If the problem persists, could you share your test program?

[Globik]

Аnd I am on windows with msys2. I cant find netinet6 with headers files. What is the name of such library for netinet6?

[tuexen]

I guess the files are in
https://github.com/sctplab/usrsctp/tree/master/usrsctplib/netinet6
You should be able to build the userland stack with nmake or cmake.

[tuexen]

OK, found the test program. Wasn't reading the screenshot... We will test it.

[tuexen]

I added a simple test program test_libmgmt.c and it doesn't show any leaks on Linux. We will test it on Windows next week.

Please note that you must use

while (usrsctp_finish() != 0) {
#ifdef _WIN32
    Sleep(1000);
#else
    sleep(1);
#endif
}

instead of

usrsctp_finish();

in your test program. If usrsctp_finish() does return a non-zero value, the stack is not cleaned up and you might leak memory when just calling usrsctp_init() again.

[jswirl]

Okay. I don't have access to my Windows machine now, but I'll test it on Monday and let you know.
Thanks for your help.

[tuexen]

I now see where the leak occurs on Windows. It doesn't occur on other platforms. Need to think about a correct way of fixing it. Will let you know when I think the issue is fixed.

[jswirl]

I can confirm that there are still leaks using the most recent code, this time I waited for usrsctp_finish() to return properly before calling usrsctp_init()

[jswirl]

@tuexen: just out of curiosity, where is the leak that you said you saw?
thanks.

[tuexen]

The allocations in
https://github.com/sctplab/usrsctp/blob/master/usrsctplib/netinet/sctp_bsd_addr.c#L357
and
https://github.com/sctplab/usrsctp/blob/master/usrsctplib/netinet/sctp_bsd_addr.c#L405
are never freed.
On the other platforms we use a global list which is freed at the end. In the FreeBSD Kernel we use pointers to a global list of addresses which never goes away.
I need to check if we actually use these pointers in the userland stack, or if a NULL pointer can be provided. If we use the global list, it should be updated when addresses are added to deleted to the stack. Not sure if this happens. I need to check that.

[tuexen]

OK. This should be fixed now. Can you retest?

[jswirl]

@tuexen: Yes, it appears to be no longer leaking. Thanks so much for your help!