As discussed, we keep the ASSERT() on "too short buffer" (because it would
be a breach of contract between *our* driver in the kernel) - but it gets
moved so even in that case we avoid the "access out of buffer memory first,
ASSERT() then" situation.
Also, code cleanup :-) - no real change, just move the common parts after
the switch().
Stared-at-code, test compiled on ubuntu/mingw. Build errors in buildbot
are unrelated "infrastructure" issues (as the unix code paths are fully
untouched by this patch).
Your patch has been applied to the master and release/2.7 branch (bugfix),
code is not in 2.6 or earlier.
commit 71b271d9a168b7c43f3b78bdc4768edadfa25e86 (master)
commit d849dbd395fff72ffc10e0161e5a4237f72f5be3 (release/2.7)
Author: Lev Stipakov
Date: Fri Jun 5 16:18:02 2026 +0200
socket: assert buffer length before reading prepended sockaddr family
Signed-off-by: Lev Stipakov <[email protected]>
Acked-by: Gert Doering <[email protected]>
Gerrit URL: https://gerrit.openvpn.net/c/openvpn/+/1706
Message-Id: <[email protected]>
URL:
https://www.mail-archive.com/[email protected]/msg37065.html
Signed-off-by: Gert Doering <[email protected]>
--
kind regards,
Gert Doering
_______________________________________________
Openvpn-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openvpn-devel
