On Wed, 27 Mar 2024 09:56:59 -0700
Tyler Retzlaff <roret...@linux.microsoft.com> wrote:
> On Wed, Mar 27, 2024 at 09:45:21AM -0700, Stephen Hemminger wrote:
> > Windows does not have localtime_r but it does have a similar
> > function that can be used instead.
> >
> > Signed-off-by: Stephen Hemminger <step...@networkplumber.org>
> > ---
> > lib/eal/windows/include/rte_os_shim.h | 10 ++++++++++
> > 1 file changed, 10 insertions(+)
> >
> > diff --git a/lib/eal/windows/include/rte_os_shim.h
> > b/lib/eal/windows/include/rte_os_shim.h
> > index eda8113662..e9741a9df2 100644
> > --- a/lib/eal/windows/include/rte_os_shim.h
> > +++ b/lib/eal/windows/include/rte_os_shim.h
> > @@ -110,4 +110,14 @@ rte_clock_gettime(clockid_t clock_id, struct timespec
> > *tp)
> > }
> > #define clock_gettime(clock_id, tp) rte_clock_gettime(clock_id, tp)
> >
> > +static inline struct tm *
> > +rte_localtime_r(const time_t *timer, struct tm *buf)
> > +{
> > + if (localtime_s(buf, timer) == 0)
> > + return buf;
> > + else
> > + return NULL;
> > +}
> > +#define localtime_r(timer, buf) rte_localtime_r(timer, buf)
>
> hm, i'm always a bit concerned about expressing platform standard names
> from dpdk api surface. i think we should just expose and use
> rte_localtime_r() and not present localtime_r.
>
> can be treated as a suggestion.
>
> Acked-by: Tyler Retzlaff <roret...@linux.microsoft.com>
>
I was just copying what clock_gettime was doing.
But yes, should really do global update to rte_localtime()
