This is the mail archive of the mailing list for the Cygwin project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: From Brent Jensen

Try using google to search the cygwin mailing list archives for 

windows error 487

I saw two different people with this same problem in the past month or 
so- I don't remember the exact solution (I think it was "check the
source code") but you should get some hits if you search for it.


"Jensen, Brent" wrote:
> Recently I got a new pc, an Athlon with 1.5 Gb sdram memory.  The operating
> system is Windows 2000.  Upon trying to run a g77 (cygwin) fortran program
> requiring about 1 Gb physical memory, I got the following error, highlighted
> in red:
> administrator@PERSEPHONE /cygdrive/c/awork1/fortran_mathcad/gd
> $ g77 blast2.f
> administrator@PERSEPHONE /cygdrive/c/awork1/fortran_mathcad/gd
> $ ./a.exe
> c:\awork1\fortran_mathcad\gd\a.exe: *** Couldn't reserve space for cygwin's
> heap
>  (0x1A020000) in child, cygheap, Win32 error 487
> administrator@PERSEPHONE /cygdrive/c/awork1/fortran_mathcad/gd
> $
> Do you know anything about this?  I ran some quick tests, by increasing an
> array size to see where the error occured. The test fortran program
> contained only a single array, a1.  When the array size was
> a1(66,1000,1000), the simple program compiled and ran.  When the a1 array
> size was increased to a1(67,1000,1000), the program would not compile,
> giving insufficient memory as the problem.  When a1 was set to
> a1(68,1000,1000) and above, I got the error highlighted in red.  a1 is a
> double precision array which I think requires 8 bytes per array element.
> This seemed to be consistent with the task manager memory usage performance
> window readings, that I had open and noted during fortran execution.
> Currently I am thinking I can possibly control heap space in C or C++, and
> may be able to execute the fortran program inside C or C++.  I am in the
> learning stages of C and C++ and currently looking into this possibility.
> Thanks for any light you might shed on this.
> --
> Unsubscribe info:
> Bug reporting:
> Documentation:
> FAQ:         

Unsubscribe info:
Bug reporting:

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]