Package: linux-image-6.6.13+bpo-amd64
Version: 6.6.13
Severity: grave
The issue was fixed already in the 6.1 kernels, see
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1057843
From that bug report I read that all kernels are safe if both commits are
present:
91562895f803 ("ext4: properly sync file size update after O_SYNC direct IO")
936e114a245b ("iomap: update ki_pos a little later in iomap_dio_complete")
This is true for the changelog of the default 6.1 kernels:
https://metadata.ftp-master.debian.org/changelogs//main/l/linux-signed-amd64/linux-signed-amd64_6.1.85+1_changelog
so in those everything is fine.
But in the changelog of the backport kernels the second line (iomap: update
ki_pos a little later in iomap_dio_complete) is missing. See:
https://metadata.ftp-master.debian.org/changelogs//main/l/linux-signed-amd64/linux-signed-amd64_6.6.13+1~bpo12+1_changelog
This is the line which fixes the issue induced by the first line.
So, if I'm not mistaken, this means that in the current debian backport kernels
linux-image-6.5.0-0.deb12.4-amd64 and linux-image-6.6.13+bpo-amd64 the ext4
data loss issue is still unfixed. Both kernels are currently distributed via
apt from the debian repos.
