Re: Possible bug retrieving IfIndex in newlib - winsup/cygwin/net.cc

[David Bean <d_bean@hotmail.com>]

Hello Corrina,

No I do not, it just stuck out while I was looking through the code hunting down another problem. I have had a long history of bumping into undocumented issues with Windows over the years and have become somewhat untrusting with the windows API. I was looking through net.cc because a samba call to if_nametoindex was failing to return an index, but it seems that Cygwin just passes that call through to the IP Help API.

Have a good one,

David Bean

________________________________
From: Corinna Vinschen <corinna-cygwin@cygwin.com>
Sent: Tuesday, October 22, 2019 2:39 PM
To: David Bean <d_bean@hotmail.com>
Cc: cygwin@cygwin.com <cygwin@cygwin.com>
Subject: Re: Possible bug retrieving IfIndex in newlib - winsup/cygwin/net.cc

Hi David,

On Oct 22 15:56, David Bean wrote:
> Good Day,
>
> I have been working on porting Samba 4.11 to Cygwin for a few days and ran into an odd issue. Samba configures its interfaces in several steps, but it relies pretty heavily on getting information from the interface structures populated by Cygwin. While I was walking through the call map I noticed a lot of the data is populated by get_ifs and that routine may have a bug in its handling of IfIndex. Basically it seems to copy the IPv4 index at all times even though it may need Ipv6Index.
>
> the same code exists at line 1904 and 1808 as well. I believe the current code may fail to retrieve the correct adapter index if IPv6 is enabled on the interface.
> this is current code:  ifp->ifa_hwdata.ifa_ifindex = pap->IfIndex;
>
> I think it should be:  ifp->ifa_hwdata.ifa_ifindex = (sa->sa_family == AF_INET
>                                          ? pap->IfIndex : pap->Ipv6Index);
>
> This is the model used by the lines retrieving the other elements with dual locations on the windows side.
>
> I am unable to debug this to confirm it, but please let me know if this is actually a bug or not.

The code in line 1808 only handles AF_INET anyway, so it shouldn't
use IpV6IfIndex.  As for line 1904, I didn't perform this distinction
because in all my testing the indices were the same.

Do you have proof that this isn't always the case?  If so, I check in
a patch.


Thanks,
Corinna

--
Corinna Vinschen
Cygwin Maintainer

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple