mbox series

[v4,0/5] y2038: Make ppoll, io_pgetevents and pselect y2038 safe

Message ID 20180920044108.6606-1-deepa.kernel@gmail.com (mailing list archive)
Headers show
Series y2038: Make ppoll, io_pgetevents and pselect y2038 safe | expand

Message

Deepa Dinamani Sept. 20, 2018, 4:41 a.m. UTC 
The series transitions the ppoll, io_getevents, and pselect syscalls
to be y2038 safe.

This is part of the work proceeding for syscalls for y2038.
It is based on the series [1] from Arnd Bergmann.

The overview of the series is as below:
1. Refactor sigmask handling logic for the above syscalls.
2. Provide y2038 safe versions of the syscalls for all ABIs.

[1] https://lkml.org/lkml/2018/8/27/651

Changes since v3:
* fixed pselect copy+paste error
Changes since v2:
* remove 64BIT_TIME conditional for ppoll, pselect,
  io_getpevents as per review comments
Changes since v1:
* fixed bug pointed out by arnd

Deepa Dinamani (5):
  signal: Add set_user_sigmask()
  signal: Add restore_user_sigmask()
  ppoll: use __kernel_timespec
  pselect6: use __kernel_timespec
  io_pgetevents: use __kernel_timespec

 fs/aio.c                 | 134 ++++++++++-----
 fs/eventpoll.c           |  52 +-----
 fs/select.c              | 360 ++++++++++++++++++++++-----------------
 include/linux/compat.h   |  20 +++
 include/linux/signal.h   |   4 +
 include/linux/syscalls.h |  20 ++-
 kernel/signal.c          |  78 +++++++++
 7 files changed, 426 insertions(+), 242 deletions(-)

Comments

Arnd Bergmann Dec. 6, 2018, 4:26 p.m. UTC | #1 
On 9/20/18, Deepa Dinamani <deepa.kernel@gmail.com> wrote:
> The series transitions the ppoll, io_getevents, and pselect syscalls
> to be y2038 safe.
>
> This is part of the work proceeding for syscalls for y2038.
> It is based on the series [1] from Arnd Bergmann.
>
> The overview of the series is as below:
> 1. Refactor sigmask handling logic for the above syscalls.
> 2. Provide y2038 safe versions of the syscalls for all ABIs.
>
> [1] https://lkml.org/lkml/2018/8/27/651

I'm sorry for dropping the ball on this earlier, so it already
missed the 4.20 merge window, and is getting almost too
late for 4.21 as well now.

I've applied your five patches to my y2038 tree now, so
it should at least be in linux-next as of tomorrow and
make it into 4.21. I'll see what other patches I have in
my backlog that we should get into that release before
I send him a pull request.

      Arnd