bug-gmp
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Can't multiply but can mpf_str_out?!


From: Ronald Bruck
Subject: Can't multiply but can mpf_str_out?!
Date: Fri, 13 Apr 2001 12:08:41 -0700

Before filing a more thorough report, has anyone had experience with situations where you can otherwise manipulate objects of type mpf_t, EXCEPT you can't multiply them?

For example, I have a couple of variables (parts of matrices) which I can print--they look completely normal; can add--but--can't multiply? Their exponent range is quite normal.

Usually these problems are my own mistakes, and I find them; an array overrunning its bounds, a variable which I init twice, or clear twice. But this one looks different.

Another question: are there problems with clearing large numbers of mpf_t's at a time? I have matrix allocation routines (basically using the mechanism of Numerical Recipes, in which you increment/decrement pointers so you don't have to use vectors [0..n] or matrices [0..n, 0..m] but can use arbitrary intervals). But when I go in to clear the mpf_t's, it crashes. And usually in such a way that backtrace shows a corrupted stack. But when I examine the memory addresses, they're where theys should be. (Again, it could be another array overrunning its bounds.)

As I said, I'll try to isolate the problem from the rest of the program and post a more formal question. But are there reports of problems with the memory allocation routines? (I give it a 95% chance the problem is mine.)

--Ron Bruck



reply via email to

[Prev in Thread] Current Thread [Next in Thread]