Re: double free corruption?

marcelo Cortez <jmdc_marcelo(at)yahoo(dot)com(dot)ar> writes:
> sorry i forgot to mention
> i'm developing one c external program, may be fault is
> my code , but surprise to me the message, what bad
> practice generate this behavior?
> fail seems to be not to reproducible all times,

> i'm using beta3 version, it's this important?

What it looks like to me is something clobbering memory, eg writing more
data into a palloc'd memory chunk than will fit; which results in
overwriting malloc's own data structures, causing malloc to complain
when it notices.

Whether it's your bug or something wrong in the beta is hard to say.

> pd: gdb can help? ( for debugging my own code)

--enable-cassert would probably help more, since it would turn on some
clobber-detection support in PG.

regards, tom lane