[dpdk-dev] [PATCH v5 5/9] app/testpmd: add clock_gettime_monotonic

[Jie Zhou]

On Sun, Apr 18, 2021 at 05:20:42PM +0000, Tal Shnaiderman wrote:
> > Subject: [dpdk-dev] [PATCH v5 5/9] app/testpmd: add
> > clock_gettime_monotonic
> > 
> > External email: Use caution opening links or attachments
> > 
> > 
> > Add clock_gettime_monotonic for testpmd on Windows
> > 
> > Signed-off-by: Jie Zhou <jizh at microsoft.com>
> > Signed-off-by: Jie Zhou <jizh at linux.microsoft.com>
> > ---
> >  app/test-pmd/config.c | 33 ++++++++++++++++++++++++++++++++-
> >  1 file changed, 32 insertions(+), 1 deletion(-)
> > 
> > diff --git a/app/test-pmd/config.c b/app/test-pmd/config.c index
> > ef0b9784d..a5f8fec5b 100644
> > --- a/app/test-pmd/config.c
> > +++ b/app/test-pmd/config.c
> > @@ -63,6 +63,12 @@
> > 
> >  #define NS_PER_SEC 1E9
> > 
> > +#ifdef RTE_EXEC_ENV_WINDOWS
> > +#define _clock_gettime_monotonic(cur_time)
> > +clock_gettime_monotonic(&cur_time)
> > +#else
> > +#define _clock_gettime_monotonic(cur_time)
> > clock_gettime(CLOCK_TYPE_ID,
> > +&cur_time) #endif
> > +
> 
> I think this change should be in EAL:
> rte_get_time_monotonic(&cur_time)
> 
> Windows implementation in windows\eal_timer.c
> Unix implementation in unix\eal_unix_timer.c

Originally and internally, the function was added into eal. But then restricted the functionality just inside testpmd to avoid currently seems unnecessary version change, per a discussion in community meeting several weeks back. If we believe eal support of clock_gettime for windows will benefit other drivers/apps now instead of future when real need comes up, I can move it back into eal. DmitryK and Tyler, any conern or inputs here?

> 
> >  static char *flowtype_to_str(uint16_t flow_type);
> > 
> >  static const struct {
> > @@ -170,6 +176,27 @@ print_ethaddr(const char *name, struct
> > rte_ether_addr *eth_addr)
> >         printf("%s%s", name, buf);
> >  }
> > 
> > +#ifdef RTE_EXEC_ENV_WINDOWS
> > +static int
> > +clock_gettime_monotonic(struct timespec *tp) {
> > +       LARGE_INTEGER pf, pc;
> > +       LONGLONG nsec;
> > +
> > +       if (QueryPerformanceFrequency(&pf) == 0)
> > +               return -1;
> > +
> > +       if (QueryPerformanceCounter(&pc) == 0)
> > +               return -1;
> > +
> > +       nsec = pc.QuadPart * NS_PER_SEC / pf.QuadPart;
> > +       tp->tv_sec = nsec / NS_PER_SEC;
> > +       tp->tv_nsec = nsec - tp->tv_sec * NS_PER_SEC;
> > +
> > +       return 0;
> > +}
> > +#endif
> > +
> >  void
> >  nic_stats_display(portid_t port_id)
> >  {
> > @@ -186,6 +213,8 @@ nic_stats_display(portid_t port_id)
> > 
> >         static const char *nic_stats_border = "########################";
> > 
> > +       int ret;
> > +
> >         if (port_id_is_invalid(port_id, ENABLED_WARN)) {
> >                 print_valid_ports();
> >                 return;
> > @@ -202,7 +231,9 @@ nic_stats_display(portid_t port_id)
> >                "%-"PRIu64"\n", stats.opackets, stats.oerrors, stats.obytes);
> > 
> >         diff_ns = 0;
> > -       if (clock_gettime(CLOCK_TYPE_ID, &cur_time) == 0) {
> > +
> > +       ret = _clock_gettime_monotonic(cur_time);
> > +       if (ret == 0) {
> >                 uint64_t ns;
> > 
> >                 ns = cur_time.tv_sec * NS_PER_SEC;
> > --
> > 2.30.0.vfs.0.2