Just answering to dan and the lists to avoid noise ^^^ : These are all false positives as non-sentinel procname gets set to a string on the register sysctl path. ^^^ : procname gets set to NULL but i
Hey Dan yes Awesome. That is homework on my part for V2. It gives me a way forward. Thx!!! Yes -- Joel Granados Attachment: signature.asc Description: PGP signature
not even. It was that the patch got mangled by some error on my side. All this applies cleanly to mcgrof's sysctl-next branch here https://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux.git/log
Yep, The thread got mangled on the way out. But hopefully the rest of the patch made its way to the lists and maintainers. Very good catch! I have fixed this as well as ensure_safe_net_sysctl that wa
I was wrong here, the patchset just hadn't all hit the mailing lists. I can't apply this patchset to anything. I tried linux-next, net, and net-next. So it's hard to review. It looks like ensure_safe
On reflecting the patch is obviously wrong. It should be pass zero as table_size in that case. See diff at the end. There is a similar bug in neigh_sysctl_register() where we use memset to zero out t
The patchset doesn't include the actual interesting changes, just a bunch of mechanical prep work. For example, in lowpan_frags_ns_sysctl_register() the sentinel is sometimes element zero if the user
In order to remove the end element from the ctl_table struct arrays, we explicitly define the size when registering the targets. The register_net_sysctl function is an indirection function that gets