On Tue, Nov 9, 2010 at 8:24 PM, Dr. David Kirkby <david.kirkby@onetel.net> wrote:
On 11/ 9/10 07:16 PM, Juan Jose Garcia-Ripoll wrote:
On Tue, Nov 9, 2010 at 7:50 PM, Dr. David Kirkby<david.kirkby@onetel.net>wrote:

On 11/ 7/10 10:00 PM, Juan Jose Garcia-Ripoll wrote:

On Sun, Nov 7, 2010 at 9:54 PM, Dr. David Kirkby<david.kirkby@onetel.net

 Unless I am mistaken, this problem is only seen with gcc. In which case,
the removal of the optimiser can be condition on __GNUC__ being defined.

Actually the optimization _only_ works with GCC, because it is the only
compiler with computed gotos.

The change does not appear to have fixed the problem - at least on a 64-bit
build on OpenSolaris using gcc 4.5.0[...]
What system did you test this on?

On two machines, fulvia and t2, one is Solaris/Sparc the other one
Solaris/Intel. I no longer have OpenSolaris.


Thank you. I will investigate. I can give you access to a fast OpenSolaris machine easy enough. But I'm puzzled why it works for you and not for me. I took the 10.4.1 source and updated that one file. I'll try with the latest git snapshot of ECL.

Maybe that is not the only change that fixed it. There have been a lot of changes between 10.4 and CVS. But perhaps you wish to wait until I settle up the testing with CVS and those machines (fulvia and t2). I mean until you see an OK here: http://ecls.sourceforge.net/logs.html


Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain)