[uClibc]Re: new "malloc" directory ...

[Miles Bader]

"Marshall M. Midden" <m4 at brecis.com> writes:
> If I use the new malloc, sometimes -- seldom, and randomly, not necessarily
> the first time runs.  Otherwise the program will display the sysDescr field,
> then a check in sys_munmap() yells that memory is being freed that was never
> allocated to the program (and it won't free it).  [and program terminates]
> 
> Any ideas?  Oh, uclibc from snapshot Sept 6th, 2002.

It's a bit hard to debug unless you have an easily reproducable test case.

One possible method might be to use a malloc wrapper that does more
consistency checking, to detect the case of the program writing
past the end of allocated memory and trashing malloc data-structures
(a common cause of malloc failure).

-Miles
-- 
Is it true that nothing can be known?  If so how do we know this?  -Woody Allen