Re: ACATS Target Requirement Question

[Laurent GUERBY <laurent at guerby dot net>]

On Thu, 2008-09-25 at 14:49 -0500, Joel Sherrill wrote:
> Well I guess running the ACATS gives a big hint.  There were not many
> failures on sh-rtems4.10 and some of those were illegal memory accesses.
> So only the c4* and cxa* are likely to be related to not having a real
> clock tick.

The c4 fails are likely real codegen bug. As for tasking tests,
most of them have very long delays (sleep) since Ada semantics
does not guarantee anything, so it's not surprising that
most test will pass RTEMS idle task trick.

Laurent

> sh-rtems4.10-gcc (GCC) 4.4.0 20080918 (experimental) [trunk revision 140475]
> 
>                 === acats Summary ===
> # of expected passes        2299
> # of unexpected failures    13
> # of unsupported tests        3
> 
> ce2108f - simulator does not have persistent files
> ce2108h - simulator does not have persistent files
> ce3112d - simulator does not have persistent files
> a22006b - unknown reason for failure
> c380004 - Memory exception at 44000 (illegal address)
> c45532m - raised CONSTRAINT_ERROR : c45532m.adb:143 range check failed
> c45532o - raised CONSTRAINT_ERROR : c45532o.adb:147 range check failed
> c45532p - raised CONSTRAINT_ERROR : c45532p.adb:141 range check failed
> c953002 - ran too long, may eventually finish simsh7032:c953002: Memory 
> exception at 44000 (illegal address)
> c974013 - Memory exception at 44000 (illegal address)
> cd1c04e - explicit fail THE VALUES OF DERIVED_TYPE COMPONENTS WERE 
> INCORRECT.
> cxa4009 - explicit fail Incorrect results
> cxa4020 - explicit fail Incorrect results
> cxg2018 - Memory exception at 44000 (illegal address)
> cxg2019 - Memory exception at 44000 (illegal address)
> cxg2021 - Memory exception at 44000 (illegal address)
> 
> 
> Joel Sherrill wrote:
> > Hi,
> >
> > RTEMS has BSPs for a couple of simulators that
> > do not have a source for clock tick interrupts.
> > We simulate the passage of time by having a special
> > IDLE task advance time.  This works well enough
> > when all tasks block but if a test depends on
> > something like timeslice expiration, that won't
> > happen.
> >
> > My questions are...
> >
> > + Is this sufficient to pass a reasonable subset
> > of the ACATS?
> >
> > + Is it known which, if any, tests would be expected
> > to fail in this environment?
> >
> > Thanks.
> >
> > --
> > Joel Sherrill, Ph.D.             Director of Research & Development
> > joel.sherrill@OARcorp.com        On-Line Applications Research
> > Ask me about RTEMS: a free RTOS  Huntsville AL 35805
> >    Support Available             (256) 722-9985
> >
> >
> >   
> 
>