On Tue, Sep 16, 2025 at 2:53 PM Jiri Olsa <jo...@kernel.org> wrote: > > If uprobe handler changes instruction pointer we still execute single > step) or emulate the original instruction and increment the (new) ip > with its length. > > This makes the new instruction pointer bogus and application will > likely crash on illegal instruction execution. > > If user decided to take execution elsewhere, it makes little sense > to execute the original instruction, so let's skip it. > > Acked-by: Oleg Nesterov <o...@redhat.com> > Acked-by: Andrii Nakryiko <and...@kernel.org> > Signed-off-by: Jiri Olsa <jo...@kernel.org> > --- > kernel/events/uprobes.c | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c > index 7ca1940607bd..2b32c32bcb77 100644 > --- a/kernel/events/uprobes.c > +++ b/kernel/events/uprobes.c > @@ -2741,6 +2741,13 @@ static void handle_swbp(struct pt_regs *regs) > > handler_chain(uprobe, regs); > > + /* > + * If user decided to take execution elsewhere, it makes little sense > + * to execute the original instruction, so let's skip it. > + */ > + if (instruction_pointer(regs) != bp_vaddr) > + goto out; > +
Peter, Ingo, Are you guys ok with us routing this through the bpf-next tree? We'll have a tiny conflict because in perf/core branch there is arch_uprobe_optimize() call added after handler_chain(), so git merge will be a bit confused, probably. But it should be trivially resolvable. > if (arch_uprobe_skip_sstep(&uprobe->arch, regs)) > goto out; > > -- > 2.51.0 >
