x86/ioperm: Fix io bitmap invalidation on Xen PV

tss_invalidate_io_bitmap() wasn't wired up properly through the pvop
machinery, so the TSS and Xen's io bitmap would get out of sync
whenever disabling a valid io bitmap.

Add a new pvop for tss_invalidate_io_bitmap() to fix it.

This is XSA-329.

Fixes: 22fe5b0439dd ("x86/ioperm: Move TSS bitmap update to exit to user work")
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/d53075590e1f91c19f8af705059d3ff99424c020.1595030016.git.luto@kernel.org

x86/ioperm: Prevent a memory leak when fork fails

In the copy_process() routine called by _do_fork(), failure to allocate
a PID (or further along in the function) will trigger an invocation to
exit_thread(). This is done to clean up from an earlier call to
copy_thread_tls(). Naturally, the child task is passed into exit_thread(),
however during the process, io_bitmap_exit() nullifies the parent's
io_bitmap rather than the child's.

As copy_thread_tls() has been called ahead of the failure, the reference
count on the calling thread's io_bitmap is incremented as we would expect.
However, io_bitmap_exit() doesn't accept any arguments, and thus assumes
it should trash the current thread's io_bitmap reference rather than the
child's. This is pretty sneaky in practice, because in all instances but
this one, exit_thread() is called with respect to the current task and
everything works out.

A determined attacker can issue an appropriate ioctl (i.e. KDENABIO) to
get a bitmap allocated, and force a clone3() syscall to fail by passing
in a zeroed clone_args structure. The kernel handles the erroneous struct
and the buggy code path is followed, and even though the parent's reference
to the io_bitmap is trashed, the child still holds a reference and thus
the structure will never be freed.

Fix this by tweaking io_bitmap_exit() and its subroutines to accept a
task_struct argument which to operate on.

Fixes: ea5f1cd7ab49 ("x86/ioperm: Remove bitmap if all permissions dropped")
Signed-off-by: Jay Lang <jaytlang@mit.edu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable#@vger.kernel.org
Link: https://lkml.kernel.org/r/20200524162742.253727-1-jaytlang@mit.edu

x86/ioperm: Add new paravirt function update_io_bitmap()

Commit 111e7b15cf10f6 ("x86/ioperm: Extend IOPL config to control ioperm()
as well") reworked the iopl syscall to use I/O bitmaps.

Unfortunately this broke Xen PV domains using that syscall as there is
currently no I/O bitmap support in PV domains.

Add I/O bitmap support via a new paravirt function update_io_bitmap which
Xen PV domains can use to update their I/O bitmaps via a hypercall.

Fixes: 111e7b15cf10f6 ("x86/ioperm: Extend IOPL config to control ioperm() as well")
Reported-by: Jan Beulich <jbeulich@suse.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Jan Beulich <jbeulich@suse.com>
Reviewed-by: Jan Beulich <jbeulich@suse.com>
Cc: <stable@vger.kernel.org> # 5.5
Link: https://lkml.kernel.org/r/20200218154712.25490-1-jgross@suse.com

x86/ioperm: Extend IOPL config to control ioperm() as well

If iopl() is disabled, then providing ioperm() does not make much sense.

Rename the config option and disable/enable both syscalls with it. Guard
the code with #ifdefs where appropriate.

Suggested-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

x86/ioperm: Share I/O bitmap if identical

The I/O bitmap is duplicated on fork. That's wasting memory and slows down
fork. There is no point to do so. As long as the bitmap is not modified it
can be shared between threads and processes.

Add a refcount and just share it on fork. If a task modifies the bitmap
then it has to do the duplication if and only if it is shared.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Andy Lutomirski <luto@kernel.org>

x86/ioperm: Remove bitmap if all permissions dropped

If ioperm() results in a bitmap with all bits set (no permissions to any
I/O port), then handling that bitmap on context switch and exit to user
mode is pointless. Drop it.

Move the bitmap exit handling to the ioport code and reuse it for both the
thread exit path and dropping it. This allows to reuse this code for the
upcoming iopl() emulation.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Andy Lutomirski <luto@kernel.org>

x86/ioperm: Move TSS bitmap update to exit to user work

There is no point to update the TSS bitmap for tasks which use I/O bitmaps
on every context switch. It's enough to update it right before exiting to
user space.

That reduces the context switch bitmap handling to invalidating the io
bitmap base offset in the TSS when the outgoing task has TIF_IO_BITMAP
set. The invaldiation is done on purpose when a task with an IO bitmap
switches out to prevent any possible leakage of an activated IO bitmap.

It also removes the requirement to update the tasks bitmap atomically in
ioperm().

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

x86/ioperm: Add bitmap sequence number

Add a globally unique sequence number which is incremented when ioperm() is
changing the I/O bitmap of a task. Store the new sequence number in the
io_bitmap structure and compare it with the sequence number of the I/O
bitmap which was last loaded on a CPU. Only update the bitmap if the
sequence is different.

That should further reduce the overhead of I/O bitmap scheduling when there
are only a few I/O bitmap users on the system.

The 64bit sequence counter is sufficient. A wraparound of the sequence
counter assuming an ioperm() call every nanosecond would require about 584
years of uptime.

Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

x86/ioperm: Move iobitmap data into a struct

No point in having all the data in thread_struct, especially as upcoming
changes add more.

Make the bitmap in the new struct accessible as array of longs and as array
of characters via a union, so both the bitmap functions and the update
logic can avoid type casts.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>