Re: Tornado 2.2, cross-compiling woes

[Andy Foster <ajf@observatorysciences.co.uk>]

Hi Richard,

Just to follow up on this, the original compilation
was using "-O3" (set in "configure/CONFIG.gnuCommon").

I tried changing this to "-O" and recompiling and it made no difference.
I then tried switching off optimization altogether by
setting:

CROSS_OPT=NO

in "configure/CONFIG_SITE" (this adds the "-g" flag to the compilation).

It then worked fine.

Many thanks for the suggestion!

Cheers,

Andy



On Mon, 1 Dec 2003, Richard Dabney wrote:

> If you are using compiler optimization greater than '-O' try it without it.
>
>
> On 1 Dec 2003 at 16:52, Andy Foster wrote:
>
> >
> > Hi All,
> >
> > I've just been experimenting with cross-compiling
> > EPICS 3.14.4 against Tornado 2.0.2 (vxWorks 5.4.2)
> > and Tornado 2.2 (vxWorks 5.5) and I wanted to share
> > the following result.
> >
> > (Note: This is cross-compiling with the Wind River supplied
> > cross-compiler from Solaris->PPC, so this is separate from the
> > Linux->PPC posting I made a few days back).
> >
> > I have the following simple database, containing a single calcout record
> > on a 1Hz scan:
> >
> > ********************************************
> > #! Generated by VisualDCT v2.4
> > #! DBDSTART
> > #! DBD("../dbd/simple.dbd")
> > #! DBDEND
> >
> >
> > record(calcout, "$(user):trigger") {
> >   field(SCAN, "1 second")
> >   field(CALC, "(A=2)?1:0")
> >   field(OOPT, "Transition To Zero")
> > }
> >
> >
> > #! Further lines contain data used by VisualDCT
> > #! View(3131,2961,1.3)
> > #! Record("$(user):trigger",2680,2440,0,0,"$(user):trigger")
> > ***********************************************
> >
> >
> > This DOES NOT work for me when compiled with T2.2
> > i.e. the VAL field of "$(user):trigger" stays equal to 1
> > no matter what value I set A equal to.
> >
> > But it does work fine when compiled against T2.0.2
> > i.e. with the gcc-2.7.2 based compiler.
> >
> >
> > For the case which does not work (Tornado 2.2), I have:
> > .../ccppc -v
> >
> > Reading specs from /usr/software/dev/packages/vxworks/tornado2.2/ppc/host/sun4-solaris2/bin/../lib/gcc-lib/powerpc-wrs-vxworks/gcc-2.96/specs
> > gcc version gcc-2.96 (2.96+) 19990621 AltiVec VxWorks 5.5
> > 65115 79101 83322 83686 84218 85823 85973 86999 87279 87638 \
> > 88306 88708 88966 90104 90120 90296 90648 90662
> >
> > (I have just patched this compiler as per Wind River's patch:
> > patch-tor2_2.ppc.031031 - in the hope that this would fix the problem.
> > I think those are SPR numbers which appear in the line above).
> >
> > Any comments and I wondered how many sites are actively using
> > Tornado 2.2 either with a 3.13 or a 3.14 based release and if
> > you have noticed any subtle problems? All a bit worrying...
> >
> > Cheers,
> >
> > Andy
> >
> > --------------------------------------------------------------------
> > Observatory Sciences Limited    Email: ajf@observatorysciences.co.uk
> > William James House             Tel: 44 - (0)1223 - 508259
> > Cowley Road                     Fax: 44 - (0)1223 - 508258
> > Cambridge, CB4 0WX, UK          http://www.observatorysciences.co.uk
> >
>
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Richard Dabney
> SNS-4 LINAC Controls
> M/S H820
> TA-53, Bldg 398 Rm 104
> Los Alamos National Laboratory
> Los Alamos, NM 87545 USA
> 505-665-7860
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
>
>

--------------------------------------------------------------------
Observatory Sciences Limited    Email: ajf@observatorysciences.co.uk
William James House             Tel: 44 - (0)1223 - 508259
Cowley Road                     Fax: 44 - (0)1223 - 508258
Cambridge, CB4 0WX, UK          http://www.observatorysciences.co.uk