hi bill and brian, thanks for your input. I did look at the debug libraries and will keep them in mind for future problems. but my problem was to find the routine that was causing the heap growth. and as the sensing tool ,I was using the memory viewer in task manager for my process. this was not very helpful as many procedures had loops of size 30 or so and the growth was was below the 1k cutoff of task manager.
I decided to look at "heap walk". my process had 15 heaps many of which had very many sub heaps. however what I need was the sum of all the heap sizes which was of course easy to do and had a 1 byte precision. this was quite useful as before and after a routine was called you found the size of the heap and could then find the difference. using this I found the routine that was causing the growth, and of course it was an operator error! in a routine used many times I declared and initialized mpf_t but did not clear it when leaving. it gave a leak of 8 bytes measured by my heapsize monitor and when I cleared it when leaving the leak was fixed!!! at this point my memory does not grow after the first loop during which time it should grow. it is noted that this leak eventually caused a memory increase of several hundred meg in my process. you may have in fact created such a tool in your debug efforts. thank you both for your consideration and help. david > > -- You received this message because you are subscribed to the Google Groups "mpir-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to mpir-devel+unsubscr...@googlegroups.com. To post to this group, send email to mpir-devel@googlegroups.com. Visit this group at http://groups.google.com/group/mpir-devel. For more options, visit https://groups.google.com/d/optout.
