On Fri, Mar 30, 2012 at 19:21, Anthony Liguori <anth...@codemonkey.ws> wrote:
> On 03/30/2012 02:19 PM, Blue Swirl wrote:
>>
>> Avoid this warning on OpenBSD:
>> CC tests/rtc-test.o
>> /src/qemu/tests/rtc-test.c: In function 'check_time':
>> /src/qemu/tests/rtc-test.c:171: warning: format '%ld' expects type
>> 'long int', but argument 2 has type 'time_t'
>> /src/qemu/tests/rtc-test.c:173: warning: format '%ld' expects type
>> 'long int', but argument 2 has type 'time_t'
>>
>> Signed-off-by: Blue Swirl<blauwir...@gmail.com>
>
>
> Reviewed-by: Anthony Liguori <aligu...@us.ibm.com>
Thanks for the review, applied.
Actually the logic may fail in 2038 if longs are still 32 bits. Better
not run qtest then.
> Regards,
>
> Anthony Liguori
>
>> ---
>>
>> v2: make t and s longs.
>>
>> ---
>> tests/rtc-test.c | 6 +++---
>> 1 files changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/tests/rtc-test.c b/tests/rtc-test.c
>> index c32b36a..983a980 100644
>> --- a/tests/rtc-test.c
>> +++ b/tests/rtc-test.c
>> @@ -161,12 +161,12 @@ static void check_time(int wiggle)
>> }
>>
>> if (!(tm_cmp(&start, datep)<= 0&& tm_cmp(datep,&end)<= 0)) {
>>
>> - time_t t, s;
>> + long t, s;
>>
>> start.tm_isdst = datep->tm_isdst;
>>
>> - t = mktime(datep);
>> - s = mktime(&start);
>> + t = (long)mktime(datep);
>> + s = (long)mktime(&start);
>> if (t< s) {
>> g_test_message("RTC is %ld second(s) behind
>> wall-clock\n", (s - t));
>> } else {
>
>
