[Bug 1748] SPARC BSPs nanoseconds skipping back when counter wraps during get TOD or uptime call

[bugzilla-daemon at rtems.org]

https://www.rtems.org/bugzilla/show_bug.cgi?id=1748

Jorge Lopez <jorge.lopez.trescastro at esa.int> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
                 CC|                            |jorge.lopez.trescastro at esa.
                   |                            |int
         Resolution|FIXED                       |

--- Comment #12 from Jorge Lopez <jorge.lopez.trescastro at esa.int> 2011-04-20 07:20:18 CDT ---
I think a line is missing in the current LEON2 implementation:

http://www.rtems.com/viewvc/rtems/c/src/lib/libbsp/sparc/leon2/clock/ckinit.c?view=markup


62    uint32_t bsp_clock_nanoseconds_since_last_tick(void)
63    {
64      uint32_t clicks;
65      uint32_t usecs;
66    

+         clicks = LEON_REG.Timer_Counter_1;

67      if ( LEON_Is_interrupt_pending( LEON_INTERRUPT_TIMER1 ) ) {
68        clicks = LEON_REG.Timer_Counter_1;
69        usecs = (2*rtems_configuration_get_microseconds_per_tick() - clicks);
70      } else {
71        usecs = (rtems_configuration_get_microseconds_per_tick() - clicks);
72      }
73      return usecs * 1000;
74    }

Otherwise, we will not read the timer1 counter if a timer1 interrupt is not
pending.

Jorge.

-- 
Configure bugmail: https://www.rtems.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.