Re: [PATCH v2 2/3] kgdb: Use the kprobe blocklist to limit single stepping

From: Daniel Thompson
Date: Fri Sep 04 2020 - 11:22:58 EST


On Tue, Jul 21, 2020 at 02:04:45PM -0700, Doug Anderson wrote:
> On Mon, Jul 20, 2020 at 1:08 AM Daniel Thompson
> <daniel.thompson@xxxxxxxxxx> wrote:
> > On Fri, Jul 17, 2020 at 03:39:51PM -0700, Doug Anderson wrote:
> > > On Thu, Jul 16, 2020 at 8:20 AM Daniel Thompson
> > > <daniel.thompson@xxxxxxxxxx> wrote:
> > > >
> > > > If we are running in a part of the kernel that dislikes breakpoint
> > > > debugging then it is very unlikely to be safe to single step. Add
> > > > some safety rails to prevent stepping through anything on the kprobe
> > > > blocklist.
> > > >
> > > > As part of this kdb_ss() will no longer set the DOING_SS flags when it
> > > > requests a step. This is safe because this flag is already redundant,
> > > > returning KDB_CMD_SS is all that is needed to request a step (and this
> > > > saves us from having to unset the flag if the safety check fails).
> > > >
> > > > Signed-off-by: Daniel Thompson <daniel.thompson@xxxxxxxxxx>
> > > > ---
> > > > include/linux/kgdb.h | 1 +
> > > > kernel/debug/debug_core.c | 13 +++++++++++++
> > > > kernel/debug/gdbstub.c | 10 +++++++++-
> > > > kernel/debug/kdb/kdb_bp.c | 8 ++------
> > > > kernel/debug/kdb/kdb_main.c | 10 ++++++++--
> > > > 5 files changed, 33 insertions(+), 9 deletions(-)
> > > >
> > > > diff --git a/include/linux/kgdb.h b/include/linux/kgdb.h
> > > > index 7caba4604edc..aefe823998cb 100644
> > > > --- a/include/linux/kgdb.h
> > > > +++ b/include/linux/kgdb.h
> > > > @@ -214,6 +214,7 @@ extern void kgdb_arch_set_pc(struct pt_regs *regs, unsigned long pc);
> > > >
> > > > /* Optional functions. */
> > > > extern int kgdb_validate_break_address(unsigned long addr);
> > > > +extern int kgdb_validate_single_step_address(unsigned long addr);
> > > > extern int kgdb_arch_set_breakpoint(struct kgdb_bkpt *bpt);
> > > > extern int kgdb_arch_remove_breakpoint(struct kgdb_bkpt *bpt);
> > > >
> > > > diff --git a/kernel/debug/debug_core.c b/kernel/debug/debug_core.c
> > > > index 133a361578dc..4b59bcc90c5d 100644
> > > > --- a/kernel/debug/debug_core.c
> > > > +++ b/kernel/debug/debug_core.c
> > > > @@ -208,6 +208,19 @@ int __weak kgdb_validate_break_address(unsigned long addr)
> > > > return err;
> > > > }
> > > >
> > > > +int __weak kgdb_validate_single_step_address(unsigned long addr)
> > > > +{
> > > > + /*
> > > > + * Disallow stepping when we are executing code that is marked
> > > > + * as unsuitable for breakpointing... stepping won't be safe
> > > > + * either!
> > > > + */
> > > > + if (kgdb_within_blocklist(addr))
> > > > + return -EINVAL;
> > > > +
> > > > + return 0;
> > > > +}
> > > > +
> > > > unsigned long __weak kgdb_arch_pc(int exception, struct pt_regs *regs)
> > > > {
> > > > return instruction_pointer(regs);
> > > > diff --git a/kernel/debug/gdbstub.c b/kernel/debug/gdbstub.c
> > > > index 61774aec46b4..f1c88007cc2b 100644
> > > > --- a/kernel/debug/gdbstub.c
> > > > +++ b/kernel/debug/gdbstub.c
> > > > @@ -1041,8 +1041,16 @@ int gdb_serial_stub(struct kgdb_state *ks)
> > > > if (tmp == 0)
> > > > break;
> > > > /* Fall through - on tmp < 0 */
> > > > - case 'c': /* Continue packet */
> > > > case 's': /* Single step packet */
> > > > + error = kgdb_validate_single_step_address(
> > > > + kgdb_arch_pc(ks->ex_vector,
> > > > + ks->linux_regs));
> > >
> > > I'm a little confused. Isn't this like saying "if
> > > (i_am_standing_in_acid) dont_step_into_acid"?
> >
> > I describe it more as:
> >
> > if (we_know_there_is_acid_nearby)
> > dont_step_forward
> >
> > It is possible we are currently stepping in acid but it is also possible
> > (and reasonably likely) that we haven't stepped in it yet but will do so
> > soon.
> >
> >
> > > Specifically you're checking the _current_ PC to see if it's in the
> > > blocklist, right? ...but you've already (effectively) dropped into
> > > the debugger at that location, so if it really was a problem wouldn't
> > > we already be in trouble?
> >
> > The basic use case is where someone is stepping and we reach a PC that
> > would be blocked for a breakpoint. This will typically be due (although
> > I think it does generalize) to a function call and the safety rail will
> > be reached after we have jumped to the blocked function but before we
> > actually execute any instructions within it.
> >
> > Or putting it another way, there is no reason to worry if we start
> > somewhere "safe" and start stepping towards something on the blocklist.
> > We won't melt our shoes!
>
> I guess I still don't totally get it. So let's say we have:
>
> void dont_trace_this(...)
> {
> thing_not_to_trace_1();
> thing_not_to_trace_2();
> don_t_trace = this;
> }
> NOKPROBE_SYMBOL(dont_trace_this);
>
> void trace_me()
> {
> sing();
> dance();
> dont_trace_this();
> party();
> }
>
> So presumably the dont_trace_this() function is marked as
> NOKPROBE_SYMBOL because it's called by the kprobe handling code or by
> kgdb, right? So if we had a breakpoint there then we'd just have
> infinite recursion. Thus we want to prevent putting breakpoints
> anywhere in this function. Even though dont_trace_this() is also
> called from the trace_me() function it doesn't matter--we still can't
> put breakpoints in it because it would cause problems with the
> debugger.
>
> Now, I guess the question is: why exactly do we need to prevent single
> stepping in dont_trace_this(). In the case above where
> dont_trace_this() is called from trace_me() it would actually be OK to
> single step it, right? ...unless this is on a CPU that doesn't have a
> "single step mode" and has to implement stepping by breakpoints, of
> course.

I think you are persuading me.

Although I can think of plenty of places where it isn't safe to step I'm
struggling to think of any way for us to end up stopped in the debugger
in those places and certainly now without setting the catastrophic
(which is the only safety rail currently extant).

That means I don't think I can put up a strong enough case that this
patch is better than doing nothing!

I'll drop it for now.

> ...but maybe I'm confused

I think on the whole you've expressed things more lucidly than I have!
Nevertheless...

> and there is a reason that we shouldn't
> allow single stepping into dont_trace_this() when called from
> trace_me(). If that is the case, I'm wondering why it's OK to step
> and stop on the first instruction of the function but it's not OK to
> step and stop through the other instructions in the function.

... when we stop on the first instruction of a function then we have not
actually executed any part of it. In other words we haven't executed
anything on the blocklist.

Of course the whole issue is moot for now.


Daniel.