On Mon, 8 Jan 2018, Max R. P. Grossmann wrote: > If cpu_timer_sample_group returns -EINVAL, it will not have written into > *sample. Checking for cpu_timer_sample_group's return value precludes the > potential use of an uninitialized value of now in the following block. > Given an invalid clock_idx, the previous code could otherwise overwrite > *oldval in an undefined manner. This is now prevented. We also exploit > short-circuiting of && to sample the timer only if the result will > actually be used to update *oldval.
Sure the explanation makes sense, though you should have mentioned that _ALL_ current users call this function with a valid clockid. I'll amend the changelog. > Signed-off-by: Max R. P. Grossmann <[email protected]> > --- > kernel/time/posix-cpu-timers.c | 3 +-- > 1 file changed, 1 insertion(+), 2 deletions(-) > > diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c > index 1f27887aa194..e54638be6e19 100644 > --- a/kernel/time/posix-cpu-timers.c > +++ b/kernel/time/posix-cpu-timers.c > @@ -1189,9 +1189,8 @@ void set_process_cpu_timer(struct task_struct *tsk, > unsigned int clock_idx, > u64 now; > > WARN_ON_ONCE(clock_idx == CPUCLOCK_SCHED); > - cpu_timer_sample_group(clock_idx, tsk, &now); > > - if (oldval) { > + if (oldval && cpu_timer_sample_group(clock_idx, tsk, &now) != -EINVAL) { > /* > * We are setting itimer. The *oldval is absolute and we update > * it to be relative, *newval argument is relative and we update > -- > 2.15.1 > >
