Bug 215105
Summary: | Kernel leaks memory in ip6_dst_cache when suppress_prefix is present in ipv6 routing rules and a `fib` rule is present in ipv6 nftables rules | ||
---|---|---|---|
Product: | Networking | Reporter: | qtmlabs |
Component: | IPV6 | Assignee: | Hideaki YOSHIFUJI (yoshfuji) |
Status: | RESOLVED CODE_FIX | ||
Severity: | high | CC: | Jason, qtmlabs |
Priority: | P1 | Keywords: | trivial |
Hardware: | All | ||
OS: | Linux | ||
Kernel Version: | 5.4-rc1 and later | Subsystem: | |
Regression: | Yes | Bisected commit-id: |
Description
qtmlabs
2021-11-22 15:10:46 UTC
Can you post the patch with your analysis message in the body of it to netdev@vger.kernel.org, and include a proper Signed-off-by line on the patch? That will garner the eyeballs of a lot more people and will result in this being fixed most likely. A Signed-off-by line requires disclosing your legal name, which I am not willing to disclose. However the patch is trivial and uncopyrightable, and is the only (obvious) way to fix it. What's the recommended course of action in this case? Oh, shucks, okay. This really does look like a worthwhile patch with valuable analysis. I'll ask people more knowledgeable about process stuff what to do here. Alright, posted it on your behalf: https://lore.kernel.org/netdev/20211123124832.15419-1-Jason@zx2c4.com/ If you want to tweak this or send different revisions, I'll sign off and remail whatever you need. Nobody really chimed in after I mentioned that it should be reviewed, but it got merged anyway, so: https://git.kernel.org/pub/scm/linux/kernel/git/netdev/net.git/commit/?id=cdef485217d30382f3bf6448c54b4401648fe3f1 You can probably close this bug report and mark it as resolved. |