On Wed, Nov 29, 2017 at 12:17 AM, Deepa Dinamani [off-list ref] wrote:

On Tue, Nov 28, 2017 at 6:17 AM, Arnd Bergmann [off-list ref] wrote:

quoted

On Mon, Nov 27, 2017 at 11:29 PM, Deepa Dinamani [off-list ref] wrote:

Right. There are three options:

1. Use two configs to identify which syscalls need not be supported by
new architectures.
In this case it makes sense to say LEGACY_TIME_SYSCALLS and
COMPAT_32BIT_TIME both need to be disabled for new architectures. And,
I can reword the config to what you mention below.

2. Make the LEGACY_TIME_SYSCALLS eliminate non y2038 safe syscalls
mentioned below only.
In this case only the native and compat functions of the below
mentioned syscalls need to be identified by the config. I like this
option as this clearly identifies which syscalls are deprecated and do
not have a 64 bit counterpart. Not all architectures need to support
turning this off.

3. If we don't need either 1 or 2, then we could stick with what we
have today in the series as CONFIG_64BIT_TIME will be deleted and they
only need #ifdef CONFIG_64BIT.

Let me know if anyone prefers something else.

I think I prefer to have both LEGACY_TIME_SYSCALLS to guard
the native deprecated syscalls (disabled on 32-bit architectures after
the conversion, and enabled on 64-bit architectures until
we merge the next one), and COMPAT_32BIT_TIME to guard the
compat versions of both the deprecated and the non-deprecated
syscalls (enabled on all existing 32-bit architectures after the
conversion, and on 64-bit architectures if they provide a compat
mode for the former).

Those two are not symmetric, but I think those are the most
common combinations, and the Kconfig symbol helps document
what they are.

There is one more category for things like io_getevents() and
rt_sigtimedwait that also need two separate compat versions,
one for 32-bit time_t and one for 64-bit time_t, but it seems better
to deal with those case-by-case rather than introducing another
Kconfig symbol.

        Arnd