On Sun, Mar 19, 2017 at 4:57 AM, Deepa Dinamani deepa.kernel@gmail.com wrote:
From: Deepa Dinamani deepadinamani@google.com
struct timespec is not y2038 safe. The plan is to get rid of all uses of timespec internally in the kernel. Replace uses of timespec with timespec64. The syscall interfaces themselves will be changed in a separate series.
Signed-off-by: Deepa Dinamani deepa.kernel@gmail.com
It's worth pointing out here that this patch only does half the job:
index f608941..97a883a 100644 --- a/include/linux/posix-timers.h +++ b/include/linux/posix-timers.h @@ -94,7 +94,7 @@ struct k_clock { int (*clock_adj) (const clockid_t which_clock, struct timex *tx); int (*timer_create) (struct k_itimer *timer); int (*nsleep) (const clockid_t which_clock, int flags,
struct timespec *, struct timespec __user *);
struct timespec64 *, struct timespec __user *); long (*nsleep_restart) (struct restart_block *restart_block);
You change one of the two arguments, but not the second one or the code in the restart handler that uses that __user pointer.
Your patch is a good step in the right direction, and the second half of it is definitely complicated enough to be done in a separate patch, so I think it's good to keep them separate, just add explain why this is done one at a time.
Arnd