Re: patch for printf memory use tests

[Bruno Haible]

Hi Bruce,

> This patch does a few things:
> 
> 1. It increases the number of iterations of the test-fprintf-posix3
>    and test-dprintf-posix2 to 1,500 so that 15MB of address space
>    is allowed.

This may be the right thing to do. Or maybe the right thing to do
is to getrlimit (RLIMIT_AS,) with a value that is computed as
(size of address space of current process before any malloc) + 10 MB.

In order to decide among these two choices (or possibly find other,
better ways of dealing with the problem), I need to know where the
remaining 8.9 MB of address space are allocated. That's why I asked
you already in [1]:
Could you please show us the contents of the /proc/$pid/maps file, at
a time point before the dprintf() or malloc() calls?

Bruno

[1] http://lists.gnu.org/archive/html/bug-gnulib/2011-01/msg00169.html