V2 [PATCH] nptl: Zero-extend arguments to SETXID syscalls [BZ #26248]

[H.J. Lu]

On Fri, Jul 17, 2020 at 3:27 PM Aurelien Jarno <aurelien@aurel32.net> wrote:
>
> On 2020-07-17 12:42, H.J. Lu via Libc-alpha wrote:
> > On Fri, Jul 17, 2020 at 8:02 AM Carlos O'Donell <carlos@redhat.com> wrote:
> > >
> > > On 7/16/20 8:46 AM, H.J. Lu via Libc-alpha wrote:
> > > > On Thu, Jul 16, 2020 at 5:03 AM Florian Weimer <fweimer@redhat.com> wrote:
> > > >>
> > > >> * H. J. Lu via Libc-alpha:
> > > >>
> > > >>> nptl has
> > > >>>
> > > >>> /* Opcodes and data types for communication with the signal handler to
> > > >>>    change user/group IDs.  */
> > > >>> struct xid_command
> > > >>> {
> > > >>>   int syscall_no;
> > > >>>   long int id[3];
> > > >>>   volatile int cntr;
> > > >>>   volatile int error;
> > > >>> };
> > > >>>
> > > >>>  /* This must be last, otherwise the current thread might not have
> > > >>>      permissions to send SIGSETXID syscall to the other threads.  */
> > > >>>   result = INTERNAL_SYSCALL_NCS (cmdp->syscall_no, 3,
> > > >>>                                  cmdp->id[0], cmdp->id[1], cmdp->id[2]);
> > > >>>
> > > >>> But the second argument of setgroups syscal is a pointer:
> > > >>>
> > > >>>        int setgroups(size_t size, const gid_t *list);
> > > >>>
> > > >>> But on x32, pointers passed to syscall must have pointer type so that they
> > > >>> will be zero-extended.
> > > >>>
> > > >>> Add <setxid-internal.h> to define INTERNAL_SETXID_SYSCALL_NCS and use it,
> > > >>> instead of INTERNAL_SYSCALL_NCS, for SETXID syscalls.  X32 override it
> > > >>> with pointer type for setgroups.  A testcase is added and setgroups
> > > >>> returned with EFAULT when running as root without the fix.
> > > >>
> > > >> Isn't it sufficient to change the type of id to unsigned long int[3]?
> > > >> The UID arguments are unsigned on the kernel side, so no sign extension
> > > >> is required.
> > > >>
> > > >
> > > > It works.  Here is the updated patch.  OK for master?
> > > >
> > > > Thanks.
> > > >
> > >
> > > This test should run in a container, and it should attempt two setgroups
> > > calls, one with groups and one empty with a bad address.
> >
> > Fixed.
> >
> > > > From 2af9e56c2306dc9d80a4476fa5b154a26a935557 Mon Sep 17 00:00:00 2001
> > > > From: "H.J. Lu" <hjl.tools@gmail.com>
> > > > Date: Thu, 16 Jul 2020 03:37:10 -0700
> > > > Subject: [PATCH] nptl: Zero-extend arguments to SETXID syscalls [BZ #26248]
> > > >
> > > > nptl has
> > > >
> > > > /* Opcodes and data types for communication with the signal handler to
> > > >    change user/group IDs.  */
> > > > struct xid_command
> > > > {
> > > >   int syscall_no;
> > > >   long int id[3];
> > > >   volatile int cntr;
> > > >   volatile int error;
> > > > };
> > > >
> > > >  /* This must be last, otherwise the current thread might not have
> > > >      permissions to send SIGSETXID syscall to the other threads.  */
> > > >   result = INTERNAL_SYSCALL_NCS (cmdp->syscall_no, 3,
> > > >                                  cmdp->id[0], cmdp->id[1], cmdp->id[2]);
> > > >
> > > > But the second argument of setgroups syscal is a pointer:
> > > >
> > > >        int setgroups(size_t size, const gid_t *list);
> > > >
> > > > But on x32, pointers passed to syscall must have pointer type so that they
> > > > will be zero-extended.  Since the XID arguments are unsigned on the kernel
> > > > side, so no sign extension is required.  Change xid_command to
> > > >
> > > > struct xid_command
> > > > {
> > > >   int syscall_no;
> > > >   unsigned long int id[3];
> > > >   volatile int cntr;
> > > >   volatile int error;
> > > > };
> > > >
> > > > so that all arguments zero-extended.  A testcase is added for x32 and
> > > > setgroups returned with EFAULT when running as root without the fix.
> > > > ---
> > > >  nptl/descr.h                                  |  8 ++-
> > > >  sysdeps/unix/sysv/linux/x86_64/x32/Makefile   |  4 ++
> > > >  .../sysv/linux/x86_64/x32/tst-setgroups.c     | 67 +++++++++++++++++++
> > > >  3 files changed, 78 insertions(+), 1 deletion(-)
> > > >  create mode 100644 sysdeps/unix/sysv/linux/x86_64/x32/tst-setgroups.c
>
> Is there a reason to limit that test to x32? We do not have any other
> getgroups or setgroups test in the GLIBC tree, so that simple test might
> already be able to catch issues. In addition such a test would benefit
> the other ILP32 architectures supported by GLIBC.
>

I can move it to nptl.

-- 
H.J.