On Sun, Jun 24, 2018 at 6:36 PM, Bruno Haible <br...@clisp.org> wrote:
> Jim Meyering wrote:
>> diff --git a/lib/parse-datetime.y b/lib/parse-datetime.y
>> index f14bb31dc..3d8666a4d 100644
>> --- a/lib/parse-datetime.y
>> +++ b/lib/parse-datetime.y
>> @@ -1766,6 +1766,11 @@ parse_datetime2 (struct timespec *result, char const
>> *p,
>>
>> timezone_t tz = tzdefault;
>>
>> + /* Store a local copy prior to first "goto". Without this, a prior use
>> + below of RELATIVE_TIME_0 on the RHS might translate to an assignment-
>> + to-temporary, which would trigger a -Wjump-misses-init warning. */
>> + static const relative_time rel_time_0 = RELATIVE_TIME_0;
>> +
>> if (strncmp (p, "TZ=\"", 4) == 0)
>> {
>> char const *tzbase = p + 4;
>
> With gcc 4.8.5, this produces a compilation error:
>
> CC lib/parse-datetime.o
> parse-datetime.y: In function 'parse_datetime2':
> parse-datetime.y:1772:3: error: initializer element is not constant
> make[2]: *** [lib/parse-datetime.o] Error 1
>
> Workaround: Remove the word 'static' from line 1772.
I presume that by reporting this, you think gcc-4.8.5 is still worth supporting?
FYI, even gcc-5.4 is no longer supported by upstream.
