Seeing as this is a remotely exploitable vulnerability, and routers are connected to the internet 24/7, this is probably something we want to be fixed. Is this vulnerability already gone in the latest kernel version in the master branch? What about the 18.06 and 17.01 branches? Is this also something for which we want to push out a new release so that less technically inclined people can update to protect themselves against this vulnerability? Details can be found here:

1 Like

This issue has been fixed in kernel 4.19.35, 4.14.112, 4.9.169, 4.4.179 and has been merged to OpenWrt/ master a bit over three weeks ago:;a=commitdiff;h=da5bd73d702e7cb9f7da174977b606ca12ffd96b


That's good to hear! The article mentions that the vulnerability is difficult to exploit, but should an attacker be successful the implications are pretty big. Do you think this warrants a new release so that people can update without having to build their own images?

It has also been fixed in the openwrt-18.06 branch (which will become 18.06.3) around three weeks ago as well:;a=commitdiff;h=ac3b5f00e17de2c7791748f6fa434eb93ee3b66e

lede-17.01 has not been fixed yet (4.4.167), but 17.01 is basically EOL anyways and on borrowed time.

It's safe to assume that any new version will contain the fix (as their kernels will be updated before the release again), but this doesn't require new OpenWrt releases as RDS/ RDS_TCP (the faulty features) are not enabled on any target for OpenWrt; meaning that all images prebuilt by OpenWrt are not affected (only if you build OpenWrt from source and explicitly change this configuration, you may be affected).


The features not even being enabled settles it then :slight_smile: thank you very much for your thorough answers!

This is specific to intel processors AFAIK.

No, that's an entirely different vulnerability. This one affects all processors, since it's a software bug in the Linux kernel.

1 Like

This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.