On Thursday 12 November 2015 14:12:52 Amitoj Kaur Chawla wrote:
On Wed, Nov 11, 2015 at 1:22 AM, Arnd Bergmann arnd@arndb.de wrote:
On Tuesday 10 November 2015 22:21:50 Amitoj Kaur Chawla wrote:
On Tue, Nov 10, 2015 at 10:01 PM, Arnd Bergmann arnd@arndb.de wrote:
You could also do the same thing as a an intermediate step, and do a final patch on top that removes the time_t and explains in the changelog what the impact of removing that is.
The above change would become an intermediate change, and I send a separate(final) patch using time64_t?
The final patch would remove the time_t cast, to make it build when that is no longer defined. I'm unsure about whether we want this change to be part of the kernel though, so keeping it separate gives the maintainer the choice whether to apply it or not.
I don't get a build error when I remove the time_t cast. Where am I going wrong?
I meant the reverse: if we remove the definition of time_t, we get a build error unless we also remove that cast.
At the moment, removing the time_t definition will break countless other things in the kernel, so that's hard to try out, but you can imagine what error you would get.
Arnd