[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Help-glpk] Error message with glpsol on PowerPC
From: |
Andrew Makhorin |
Subject: |
Re: [Help-glpk] Error message with glpsol on PowerPC |
Date: |
Tue, 25 Dec 2007 01:30:15 +0300 |
> Seasons greetings! I'm seeing an error message from glpsol (glpk-4.24)
> when running the example zebra.mod on a PowerPC 440EPx target platform.
> Specifically, this is what I see:
> Reading model section from zebra.mod...
> 151 lines were read
> Generating c1...
> Generating c2...
> (...I removed some lines here for brevity...)
> Generating f8...
> Generating f9...
> zebra.mod:105: drink[2.13886907407916e-231,milk] out of domain
> Model processing error
> To prove that this is not a file/kit source issue, I reconfigured the
> kit for glpsol.exe on my Windows/Cygwin (Intel) source platform. There was
> no malfunction:
> 151 lines were read
> Generating c1...
> Generating c2...
> (...I removed some lines here for brevity...)
> 5 green Japanese coffee Parliament zebra
> Model has been successfully processed
> Line 105 of zebra.mod is 'f9: drink[3,"milk"] = 1;' The fact that my
> target platform sees 3 as 2.13886907407916e-231 has me concerned that I have
> an Endian issue, or something of the like. I am cross-compiling here, so
> perhaps I'm missing a compiler directive, or something. Does anyone
> recognize what I might have done wrong?
In IEEE 754 hex of 2.13886907407916e-231 is 0x100A90A80000006D, which
does not look like 3.0 (0x4008000000000000) in any way.
Most probably some internal memory blocks used by the translator were
damaged, since several first constraints were successfully generated.
To check the package on the target platform try './glpsol --mps plan.mps'.