> On 04/14/2014 06:36 PM, 함명주 wrote: > >> MyungJoo/Kyungmin, > >> > >> Bump. Can we accept this patch please? > >> > >> -Saravana > > > > Nack. > > > > Please note that freq_table is also an optional value, which may > > be null. > > Ah, I saw that the max_freq would be zero if freq_table was NULL and I > assumed that it can't be NULL. But I see that the max_freq limit is not > applied if it's zero. Thanks for catching it. > > > Besides, please be aware that your code is under rcu_read_lock(). > > Valid point. I was just trying to keep the diff simple. No one's really > going to be catting this file often when performance matters. > > > > > > > Cheers, > > MyungJoo. > > > > ps. I'll send a related patch (avoid accessing null but not-an-error > > pointer at other sysfs nodes). Thank you for letting me catch such bugs > > anyway. > > I can go ahead and do this myself if you don't mind.
No, we don't need it. It was a false alarm. Reading again, I've found that we've already made other sysfs nodes check if either freq_table is null or its size is 0. So, we only need to look at this available_frequencies node now. I'll add some notes on the ABI doc for available_frequencies soon. Cheers, MyungJoo. > > -Saravana > > -- > The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, > hosted by The Linux Foundation > > > > > > >
