summaryrefslogtreecommitdiff
path: root/include/trace
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2023-08-31 14:52:20 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2023-08-31 14:52:20 -0700
commite7e9423db459423d3dcb367217553ad9ededadc9 (patch)
treeafa49ead99ff27e69b995d7ba99d35ceae99876f /include/trace
parentf441ff73f1ec568acef03f0ce4d5088c7e65c106 (diff)
parent5069ba84b5e67873a2dfa4bf73a24506950fa1bf (diff)
Merge tag 'v6.6-vfs.super.fixes.2' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull more superblock follow-on fixes from Christian Brauner: "This contains two more small follow-up fixes for the super work this cycle. I went through all filesystems once more and detected two minor issues that still needed fixing: - Some filesystems support mtd devices (e.g., mount -t jffs2 mtd2 /mnt). The mtd infrastructure uses the sb->s_mtd pointer to find an existing superblock. When the mtd device is put and sb->s_mtd cleared the superblock can still be found fs_supers and so this risks a use-after-free. Add a small patch that aligns mtd with what we did for regular block devices and switch keying to rely on sb->s_dev. (This was tested with mtd devices and jffs2 as xfstests doesn't support mtd devices.) - Switch nfs back to rely on kill_anon_super() so the superblock is removed from the list of active supers before sb->s_fs_info is freed" * tag 'v6.6-vfs.super.fixes.2' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: NFS: switch back to using kill_anon_super mtd: key superblock by device number fs: export sget_dev()
Diffstat (limited to 'include/trace')
0 files changed, 0 insertions, 0 deletions