Re: ioc segfault if - c seq flag is not set or +c?

[Benjamin Franksen <benjamin.franksen@helmholtz-berlin.de>]

Am Donnerstag, 14. April 2011, um 17:34:34 schrieb J. Lewis Muir:
> On 4/14/11 10:01 AM, Allison, Stephanie wrote:
> > I discovered that even building RTEMS 32 bit sequences (using
> > the RTEMS 32bit tool chain) on a linux-x86_64 build is
> > problematic as the 64bit-built SNC compiler will, in some
> > special cases, convert the .st file to a buggy .c file (in the
> > initialization of the state event masks).
>
> I had a similar problem compiling on RHEL 5 (linux-x86_64) for
> VxWorks (vxWorks-ppc603, vxWorks-ppc604, and
> vxWorks-ppc604_long) using EPICS 3.14.11 (+ Known Problems
> patches), SEQ 2.0.12 (as packaged in SynApps 5.5), and the
> Tornado 2.0.2 toolchain.  I haven't had time to investigate, but
> indeed the SNC compiler was generating .c files that the
> cross-compiler could not compile.  I think the SNC compiler was
> generating integer values greater than what could fit in 32 bits
> which the cross-compiler choked on.

Hello Lewis, Stephanie

I have a preliminary version of seq-2.0.13 that should solve the problems with
64 *target* architectures (including running the SNL program on the same
platform as the host).

However, what you both describe is another matter, namely a 64 bit host system
that generates code unsuitable for a 32 bit target system. This is difficult to
test for me because all our development machines at work are 32 bit, and on my
64 bit machine at home I don't have any cross compilers installed (nor any VME
boards etc). This means I need more detailed bug reports to fix such problems.

Anyway, the bitMask problem is easy to understand and fix even w/o a test
system (now that I know that there is a problem!) by changing

  typedef unsigned long bitMask;

(in seqCom.h) to

  typedef epicsUInt32 bitMask;

This is enough since the compiler prints it with

  printf("\t0x%08lx,\n", (unsigned long)pEventMask[n]);

(assuming that an unsigned long has at least 32 bits on all supported host
platforms, which I believe is the case).

I will include this fix in 2.0.13 which I plan to make available for testing in
the next days.

Cheers
Ben

________________________________

Helmholtz-Zentrum Berlin für Materialien und Energie GmbH

Mitglied der Hermann von Helmholtz-Gemeinschaft Deutscher Forschungszentren e.V.

Aufsichtsrat: Vorsitzender Prof. Dr. Dr. h.c. mult. Joachim Treusch, stv. Vorsitzende Dr. Beatrix Vierkorn-Rudolph
Geschäftsführer: Prof. Dr. Anke Rita Kaysser-Pyzalla, Prof. Dr. Dr. h.c. Wolfgang Eberhardt, Dr. Ulrich Breuer

Sitz Berlin, AG Charlottenburg, 89 HRB 5583

Postadresse:
Hahn-Meitner-Platz 1
D-14109 Berlin

http://www.helmholtz-berlin.de