[systemd-devel] systemd-196/196 fails to build with x32 Abi support

[Kay Sievers]

On Fri, Jan 18, 2013 at 11:11 PM, Zbigniew Jędrzejewski-Szmek
<zbyszek at in.waw.pl> wrote:
> On Fri, Jan 18, 2013 at 07:34:29PM +0100, Lennart Poettering wrote:
>> On Fri, 18.01.13 18:22, Stephan Raue (mailinglists at openelec.tv) wrote:
>>
>> >
>> > Am 18.01.2013 18:14, schrieb Colin Walters:
>> > >On Fri, 2013-01-18 at 16:14 +0100, Stephan Raue wrote:
>> > >
>> > >>src/core/manager.c: In function 'manager_setup_time_change':
>> > >>src/core/manager.c:160:14: error: static assertion failed:
>> > >>"sizeof(time_t) == sizeof(long)"
>> > >>           assert_cc(sizeof(time_t) == sizeof(long));
>> > >>                ^
>> > >You'll have to dive in here and adjust the code - this assumption is not
>> > >true on x32.
>> > >
>> > >
>> > this was clear to me, can i get any help with adjusting the code? :-)
>>
>> Not sure what the best way is to detect an architecture where time_t is
>> "long long" rather than "long". I figure there must be some macro we
>> could check for. If we have that, the code can be changed to:
>>
>> #ifdef MACRO_TO_CHECK
>>         assert_cc(sizeof(time_t) == sizeof(long long));
>>         its.it_value.tv_sec = LONG_LONG_MAX;
>> #else
>>         assert_cc(sizeof(time_t) == sizeof(long));
>>         its.it_value.tv_sec = LONG_MAX;
>> #endif
>>
>> i.e. the goal here is to initialize the time_t value that is tv_sec to
>> the maximum possible value.
> Fixed that.

Hmm, wouldn't:
  #define TIME_T_MAX (time_t)((1UL << ((sizeof(time_t) << 3) - 1)) - 1)
just do the same without all the configure magic?

Kay