Bug#780486: unblock: coreutils/8.23-4
Package: release.debian.org
Severity: norma
User: release.debian.org@packages.debian.org
Usertags: unblock
Please unblock package coreutils to address #765514. Upstream made a change to
chroot(8) in the coreutils currently in testing which skipped calling chroot(2)
if chrooting to the current root. The idea was to allow "chroot /" to always
work for any user in order to support some use cases. The change turned out to
cause problems for other use cases (specifically, when chrooting to a bind
mount of / that had different children than the original /; in that case, the
chroot(8) will appear to succeed but is actually a no-op). Upstream reverted
the change and added a new --skip-chroot optiion instead. The new debian
package does not add the new option, and only makes the chroot(2) call
unconditional. (Replaces if(!root) with if(1).) This is the same behavior as
wheezy and all previous debian releases. It does not seem like a good idea to
release with behavior already reverted upstream which may have security
implications (silently failing to call chroot(2)).
debdiff:
File lists identical (after any substitutions)
Control files: lines which differ (wdiff format)
------------------------------------------------
Installed-Size: [-13855-] {+14249+}
Version: [-8.23-3-] {+8.23-4+}
complete patch:
diff -urNad '--exclude=CVS' '--exclude=.svn' '--exclude=.git' '--exclude=.arch' '--exclude=.hg' '--exclude=_darcs' '--exclude=.bzr' coreutils-8.23~/src/chroot.c coreutils-8.23/src/chroot.c
--- coreutils-8.23~/src/chroot.c 2015-03-14 08:08:20.000000000 -0400
+++ coreutils-8.23/src/chroot.c 2015-03-14 08:09:13.219749384 -0400
@@ -271,7 +271,8 @@
/* Only do chroot specific actions if actually changing root.
The main difference here is that we don't change working dir. */
- if (! is_root (argv[optind]))
+ /* (reverted -- run unconditionally) */
+ if (1)
{
/* We have to look up users and groups twice.
- First, outside the chroot to load potentially necessary passwd/group
unblock coreutils/8.23-4
Reply to: