Re: unstable binaries

[tab@snarc.org (Vincent Hanquez)]

On Mon, Mar 07, 2005 at 09:47:20PM +0100, nils toedtmann wrote:
> Ok, after lots of compiling, rebooting & testing, i got this picture:
> 
> The error "wget: retr.c:293: calc_rate: Assertion `msecs >= 0' failed."
> is still present with xen-2.0-testing ("ChangeSet@1.1755, 2005-03-04 
> 00:57:22"), but now i have to trigger them by higher load. I was not
> able to reproduce it on a vanilla kernel with same config (besides 
> ARCH specific options).

>   <<<test_output>>>
>   nanosleep02    1  FAIL  :  Remaining sleep time 4010000 usec doesn't match with the expected 3999236 usec time
>   nanosleep02    1  FAIL  :  child process exited abnormally
>   incrementing stop
>   <<<execution_status>>>
>   duration=1 termination_type=exited termination_id=1 corefile=no
>   cutime=0 cstime=0
>   <<<test_end>>>

the failure is just a precision range problem. increasing the nanosleep
USEC_PRECISION to 10 times more make the test suceed.

Now running with HZ = 1000 make the precision problem disappear (at the
expense of awakening the test 700ns too soon and still failing ..)

Could you try running with HZ = 1000 to see if there are still problems
with your programs ?

Also the strace output of the wget program failing might be handy to
know if there's a link with nanosleep at least.

-- 
Vincent Hanquez


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click