On Fri, 26 Jul 2024 21:58:30 +0200
Mathias Krause <mini...@grsecurity.net> wrote:
> >>
> >> IMHO, this needs at least the following additional fixes tags to ensure
> >> all stable kernels get covered:
> >>
> >> Fixes: 5790b1fb3d67 ("eventfs: Remove eventfs_file and just use
> >> eventfs_inode")
> >> Fixes: 27152bceea1d ("eventfs: Move tracing/events to eventfs")
> >>
> >> Even if 27152bceea1d is not the real cause, just the commit making the
> >> bug reachable. But from looking at the history, this was always wrong?
> >
> > All stable kernels should get covered as 27152bceea1d has both a Cc
> > stable tag and a Fixes tag for 5790b1fb3d67. And the stable kernels
> > look at what commits have been backported to determine what other
> > commits should be backported.
>
> Now you lost me. Neither has 27152bceea1d a Cc stable tag, nor a Fixes
> tag for 5790b1fb3d67. It simply cannot, because it's from July 2023
> (v6.6-rc1) and 5790b1fb3d67 is from October 2024 (v6.7-rc1).
I'm juggling too many things around. I was thinking that 27152bceea1d
was b63db58e2fa5d. My mistake.
>
> > By saying this fixes 27152bceea1d, it
> > should all work out correctly.
>
> That would be fine with me, as that's what my git bisect run pointed at
> as well -- the oldest commit triggering the bug. However, in your v2
> it's still b63db58e2fa5d (which has a Fixes tag for 5790b1fb3d672 but
> not 27152bceea1d) which would suggest only kernels down to v6.7 are
> affected.
OK, I see what your saying. So the bug is present with 27152bceea1d,
but so are a lot of other bugs. This was completely rewritten with the help
from Linus, and an effort was made to backport it all to 6.6.
https://lore.kernel.org/all/20240206120905.570408...@rostedt.homelinux.com/
The above includes the 5790b1fb3d672 commit. Which is why this is the
commit I labeled as the main patch to backport to.
-- Steve
