Re: [ECOS] libc-time-clock test doesn't seem to be written correctly??

[sandeep <sandeep at codito dot com>]

> The for loop counts how long it takes for the return value of clock() to 
> change, that's all. Since the code is looped over in a consistent way, 
> how far the for loop counts should be roughly the same if you run it 
> again and again.


then following change will achieve more correctly and closely what you have 
mentioned above i.e. -- "how long it takes for the return value of clock() to 
change".

static unsigned long
clock_loop( const int timeout, clock_t prevclock, clock_t *newclock )
{
    clock_t c=0;
    long i;

// ++++++++++++++++

    c = clock();
    while ((prevclock = clock()) == c) /* do nothing */ ;

    // now when we enter the for-loop below, clock() value has just changed

// ++++++++++++++++

    for (i=0; i<timeout; i++) {
        c = clock();
        if ( c != prevclock )
            break; // Hit the next clock pulse
    }

    if (i==timeout)
        CYG_TEST_FAIL_FINISH("No change in clock state!");

// ???????????????
// why should following be a fail condition in test where we are trying to
// check for clock stability
// ???????????????

// NRN-start : Not Really Needed, in the view of what test is aiming at

    // it should not overflow in the lifetime of this test
    if (c < prevclock)
        CYG_TEST_FAIL_FINISH("Clock decremented!");

    *newclock = c;

// NRN-end

    return i;
} // clock_loop()

test possibly could be simplified more, as the clock_t arguments to clock_loop 
might not be required, if code b/w "// NRN-start ..... // NRN-end" is removed.

further leadeding to simplifications in main().


only use of obtained clock() value in main() is to test for "... meaning unimplemented", that will

remain even w/o need of clocks[].


--
regards
sandeep
-----------------------------------------------------------------------
Ashes to ashes,dust to dust,
if the Lord won't have you,
the Devil must.
-----------------------------------------------------------------------


--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss