| Message ID | 20230713141001.27046-1-cyphar@cyphar.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [v2] procfs: block chmod on /proc/thread-self/comm | expand |
On Fri, 14 Jul 2023 00:09:58 +1000, Aleksa Sarai wrote: > Due to an oversight in commit 1b3044e39a89 ("procfs: fix pthread > cross-thread naming if !PR_DUMPABLE") in switching from REG to NOD, > chmod operations on /proc/thread-self/comm were no longer blocked as > they are on almost all other procfs files. > > A very similar situation with /proc/self/environ was used to as a root > exploit a long time ago, but procfs has SB_I_NOEXEC so this is simply a > correctness issue. > > [...] Just to reiterate: The long term fix to avoid these odd behavioral bugs in the future is to remove the fallback to simple_setattr() we still have in notify_change() that happens when no setattr inode operation has been explicitly defined. To do this we need to change all filesystem that rely on this fallback to explicitly set simple_setattr() as their inode operation. Then notify_change() would simply return EOPNOTSUPP when no setattr iop is defined making such omissions pretty obvious. But that's a bigger patch. This is a backportable fix for this issue. Needs long soaking in -next ofc. --- Applied to the fs.proc.uapi branch of the vfs/vfs.git tree. Patches in the fs.proc.uapi branch should appear in linux-next soon. Please report any outstanding bugs that were missed during review in a new review to the original patch series allowing us to drop it. It's encouraged to provide Acked-bys and Reviewed-bys even though the patch has now been applied. If possible patch trailers will be updated. Note that commit hashes shown below are subject to change due to rebase, trailer updates or similar. If in doubt, please check the listed branch. tree: https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git branch: fs.proc.uapi [1/1] procfs: block chmod on /proc/thread-self/comm https://git.kernel.org/vfs/vfs/c/ccf61486fe1e
diff --git a/fs/proc/base.c b/fs/proc/base.c index 05452c3b9872..7394229816f3 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -3583,7 +3583,8 @@ static int proc_tid_comm_permission(struct mnt_idmap *idmap, } static const struct inode_operations proc_tid_comm_inode_operations = { - .permission = proc_tid_comm_permission, + .setattr = proc_setattr, + .permission = proc_tid_comm_permission, }; /*
Due to an oversight in commit 1b3044e39a89 ("procfs: fix pthread cross-thread naming if !PR_DUMPABLE") in switching from REG to NOD, chmod operations on /proc/thread-self/comm were no longer blocked as they are on almost all other procfs files. A very similar situation with /proc/self/environ was used to as a root exploit a long time ago, but procfs has SB_I_NOEXEC so this is simply a correctness issue. Ref: https://lwn.net/Articles/191954/ Ref: 6d76fa58b050 ("Don't allow chmod() on the /proc/<pid>/ files") Fixes: 1b3044e39a89 ("procfs: fix pthread cross-thread naming if !PR_DUMPABLE") Cc: stable@vger.kernel.org # v4.7+ Signed-off-by: Aleksa Sarai <cyphar@cyphar.com> --- v2: removed nolibc selftests as per review v1: <https://lore.kernel.org/linux-fsdevel/20230713121907.9693-1-cyphar@cyphar.com/> --- fs/proc/base.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)